Immediate and sustainable effects of transcranial direct current stimulation on pain reduction in older adults with Alzheimer’s disease and related dementias: a pilot study

doi:10.21203/rs.3.rs-5228344/v1

We simultaneously assessed the impact of transcranial direct current stimulation (tDCS) on mitigating both self-reported pain and pain behaviors to more objectively explore its effects in older adults with Alzheimer’s disease and related dementias. The analysis investigated 40 participants randomly (1:1) subjected to active and sham tDCS for 20 min on 5 consecutive days. Multi-group latent transition analysis enabled the simultaneous evaluation of both pain domains in a single model and analysis of their changes as a function of intervention exposure by modeling the transition probabilities of latent classes and comparing these changes between groups. Two pain categories (“high pain” and “low pain”) were identified based on the numeric rating scale and mobilization–observation–behavior–intensity–dementia scale scores. Overall, tDCS demonstrated better effects in helping participants transition to a “low pain” status during and after the intervention (~3 months) compared with sham stimulation, demonstrating its immediate and enduring effects.

Alzheimer’s disease and related dementias

community health

latent transition analysis

pain

transcranial direct current stimulation

Chronic pain is a significant health issue among community-dwelling older adults with Alzheimer’s disease and related dementias (ADRD), impacting more than 50% of these individuals.¹ If not appropriately treated, pain potentially results in several adverse outcomes, such as cognitive and physical deterioration,^2-4 reduced quality of life,^5,6 increased hospitalizations,⁷ and mortality.^8,9 Untreated pain can also exacerbate the neuropsychiatric symptoms of ADRD, such as aggression, agitation, hallucinations, and delusions.^10-12

Pharmacological treatments including opioids are frequently used to manage chronic pain in older adults with ADRD, but they often result in significant side effects and limited effectiveness.¹³ Non-invasive brain stimulation techniques, particularly transcranial direct current stimulation (tDCS), have gained widespread recognition as a promising nonpharmacological neuromodulatory approach toward improving chronic pain in older adults.^14,15 Shifting from traditional clinical settings, home-based, remotely supervised tDCS has recently emerged as a safe and viable alternative for pain therapy.^16,17 Considering its feasibility and acceptability, it certainly extends its benefits to those with ADRD who may find attending clinic-based sessions challenging. Nevertheless, a significant research gap regarding the effectiveness of home-based, remotely supervised tDCS in reducing pain in older adults with ADRD remains.

Previous studies have generally evaluated tDCS’s effect based on single, self-reported pain measures, such as the numeric rating scale (NRS).¹⁸ The pain NRS is a subjective report on daily pain experiences, reflecting both the somatosensory and emotional dimensions of pain, and it remains the “gold standard” for quantifying clinical pain intensity.¹⁹ Notwithstanding, in older adults, especially those with ADRD where cognitive decline may compromise the accuracy of self-reported pain, behavioral pain indicators—such as facial expressions, verbalizations, vocalizations, and body movements, assessed via external observations—may complement and validate self-reports. Simultaneously leveraging both subjective and objective measures may improve the reliability of pain evaluations and help enhance our understanding of older adults’ pain experience.

Therefore, we simultaneously assessed the impact of tDCS on mitigating both self-reported pain and caregiver-observed pain behaviors to more objectively and comprehensively explore its effects in older adults with ADRD using a unified analysis approach: latent transition analysis (LTA).²⁰ LTA is a person-centered approach that uses longitudinal data to handle transitions across latent classes of individuals over time.²⁰ In LTA, these latent classes are considered dynamic “statuses” rather than stable classifications, that people may move in and out of over time (e.g., movements between different pain profile groups); such movement between latent statuses is quantified in a matrix of transition probabilities.²⁰ In clinical trials, researchers can test intergroup differences (e.g., intervention vs. control) in transition probabilities over time by applying multi-group LTA, and these may modulate interventional efforts for symptom improvement.²¹ Importantly, compared with traditional approaches such as growth-curve modeling and repeated-measures analysis of variance, LTA is advantageous in that it entails the simultaneous measurement of multiple indicators (e.g., LTA evaluates the contributions of different pain variables to each latent status). Therefore, it is especially useful in capturing multifaceted pain changes across several pain-related measures to observe the intervention effect more objectively and holistically.

Study aim

We simultaneously evaluated the effects of home-based, remotely supervised tDCS on self-reported pain and caregiver-observed pain behaviors in older adults with ADRD via multi-group LTA, thereby enabling a more objective and comprehensive evaluation of intervention effects. Given the significant concern of under-reporting and under-treatment of pain in this population,²² and according to the national goal of enhancing community-based care, our study holds considerable relevance.

Design

This is a secondary analysis of a double-blind, randomized, sham-controlled, phase II, parallel-group pilot clinical trial. The trial enrolled 40 community-dwelling older adults with early-stage ADRD and divided them equally into two groups: active and sham tDCS (both n = 20) (Figure 1). An allocation sequence was generated using a randomization list formulated by a statistician uninvolved in the trial’s clinical aspects and based on the order of study enrollment. Randomization ensured balance between the two groups in terms of age, race, sex, and dementia severity. The parent trial was registered at ClinicalTrials.gov (blinded for review). Further details are available in the original study (blinded for review). Ethical approval was obtained from the participating university (blinded for review).

Participants

Individuals aged 50 to 90 with early-stage ADRD were eligible for inclusion in the study if they (1) reported chronic pain over the past three months averaging ≥ 3 on a 0-10 NRS, (2) had a caregiver who interacted with them for at least 10 hours a week, (3) could speak and read English, and (4) had no plans to change their medication regimens during the trial. A study physician confirmed the diagnosis of early-stage ADRD using the Clinical Dementia Rating (0.5 to 1.0), Mini-Mental Status Exam (16 to 23), or the telephone version of the Montreal Cognitive Assessment (16 to 26). The participants were excluded if they had medical conditions that could affect outcome interpretation, pose safety risks during assessments or tDCS procedures, or prevent protocol completion. The exclusion criteria were (1) any history of significant neurological issues (brain surgery, tumor, seizure, stroke, or intracranial metal); (2) alcohol or substance use disorders; (3) severely reduced cognitive function (Mini-Mental Status Exam score ≤ 15); and (4) hospitalization for neuropsychiatric conditions in the past year.

The eligible participants and their caregivers were scheduled for a baseline visit 3 to 7 days before the tDCS intervention. During the visit, we obtained written informed consent, assessed clinical pain and pain-related cortical responses, and provided training on home-based tDCS use. Following this, the participants were randomly divided into active or sham groups. After the 5-day intervention period, we evaluated post-intervention outcomes and collected the tDCS devices.

Ethics approval and consent to participate

The protocol has been registered at www.clinicaltrials.gov (blind for review). Ethical approval was obtained from the participating university (blind for review). Informed and written consent was obtained from all subjects involved in the study.

Intervention

Active tDCS. The home-based tDCS device was a “Soterix 1×1 tDCS mini-CT Stimulator” (Soterix Medical Inc., NY) equipped with headgear and 5×7-cm saline-soaked surface sponge electrodes. The device was administered daily for 20 min per session. A constant 2-mA current was applied, representing a standard intensity reported for its analgesic effects and previously employed in our and other studies.^16,23-25 The anode was positioned over the primary motor cortex of the left hemisphere and the cathode over the right supraorbital area, considering that this method potentially alters brain activity in a non-invasive, painless, and safe manner.²⁶ The sponge electrodes were attached to the custom headgear, which was fixed onto the participant’s head to ensure simple and foolproof electrode preparation.²⁷ Strictly after having received a unique unlock code from the research team, participants or caregivers administered each stimulation session. Once suitable contact quality had been achieved, they could only operate the on/off button and were unable to alter the device settings. The tDCS device sensed contact between the scalp and SnapPad® and indicated whether it was poor, moderate, or good. After 20 min, the device automatically switched off, and study staff instructed participants to remove and dispose of the sponges and store the equipment securely for the subsequent session. Consistency and supervision were maintained by having participants use the device at a predetermined time each weekday while seated quietly in a chair.

Sham tDCS. For sham stimulation, the setup mirrored that of the active one; however, the stimulator was only activated for 30 s at the beginning and end of the session to replicate the sensory experience of active tDCS without delivering a sustained current, effectively concealing whether the stimulation was active or sham. This method has been validated as reliable and indistinguishable from active tDCS.^28,29 All participants were informed that they may or may not feel any sensations during the intervention. The information and instructions presented on the device were identical for both the active and sham conditions.

Measurement

The collected demographic information included age, gender, body mass index (BMI; kg/m²), race, marital status, and education. The pain was assessed using the NRS for self-reported pain and the mobilization–observation–behavior–intensity–dementia (MOBID-2) scale for pain intensity rating of observed pain behavior. We utilized pain measurements taken at baseline, immediately after the 5-day intervention completion, and at one- and three-month follow-ups.

For the NRS, the participants were asked to choose a number between 0 and 100 to reflect their pain intensity, with 100 indicating maximum pain. The NRS demonstrates good reliability and validity for pain assessment in dementia patients, maintaining high internal consistency with a Cronbach’s alpha coefficient of 0.80. It is also shown to accurately capture self-reported pain in individuals with mild to moderate dementia.³⁰

The MOBID-2 scale, a validated tool effective in detecting changes in pain among individuals with ADRD, was used with caregivers. The MOBID-2 has a reported Cronbach’s alpha coefficient of ≥ .8, indicating strong reliability for detecting pain changes in individuals with ADRD.^31,32 The scale has two parts: the first part assesses nociceptive, musculoskeletal pain through five actively guided movements, during which the raters (i.e., caregivers) are encouraged to look for pain behavior; the second part, consisting of five items, evaluates pain from the head, skin, and internal organs using an NRS from 0 to 10. After these assessments, the raters compiled the results to give an overall pain score on an NRS from 0 to 10, which was used for the analyses.

Statistical analysis

Descriptive statistics were used to characterize the study participants. The Chi-square or Fisher’s exact test for categorical variables and the t-test for continuous variables were used to compare participant characteristics between the groups. Our main goal was to use multi-group LTA to investigate if the changes in both self-reported pain and caregiver-observed pain behaviors over time differed between the active and sham groups. The LTA was performed with Mplus version 8.8. Supplemental Table 1 includes the Mplus syntax.

Latent Transition Analysis (LTA). LTA is a type of structural equation model used to model transitions from one latent status to another over time.²⁰ LTA yields three sets of parameters: 1) a matrix of conditional (status-specific) item-response probabilities for each of the indicators in the measurement model at each point in time (“ρ” parameters), 2) a vector of latent status probabilities at Time 1 (“δ” parameters) describing the time-specific prevalence of each latent status, and 3) matrices of transition probabilities (“τ” parameters) representing the probability of membership in a status at timepoint t dependent upon membership in a latent status at timepoint t−1.

In multi-group LTA, latent status probabilities (δ’s) and transition (τ’s) probabilities are expressed as a function of a grouping variable, enabling between-group comparisons of the prevalence of the latent statuses and incidence of transitions over time.³³ In the context of clinical trials, this approach allows researchers to statistically assess intervention effects.³³ Theoretically, LTA parameters potentially vary between groups. For consistency in interpretation, constraining each element of the matrix of ρ parameters at Time 1 to be equal to the corresponding element at subsequent times is advisable to ensure that status definitions at each time point remain consistent, thereby ensuring measurement invariance.³⁴ This facilitates intergroup comparisons of both latent status and transition probabilities.³⁴

Main Statistical Analysis. Latent statuses were based on the participants’ responses to the NRS and MOBID-2 instruments. We initially estimated a series of unconditional LTA models with increasing numbers of latent statuses and selected the optimal model using the Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), sample-size adjusted BIC (SABIC), log-likelihood (LL), and entropy.³⁵ Where fit statistics yielded contradictory information, we checked for interpretability and clinical meaningfulness. After determining the optimal LTA model, we fitted a multi-group LTA model using the intervention condition as a grouping variable. We imposed measurement invariance to maintain consistent meanings of latent statuses across groups and time points. All analyses were conducted with all available data points using the Full Information Maximum Likelihood estimation,³⁶ which can account for indicator-level and longitudinal-level missingness.

The mean age was 71.9 ± 8.97 years in the active tDCS group and 74.2 ± 6.16 years in the sham tDCS group, with similar proportions of females (75.0%) in both groups. BMI values were 26.7 ± 5.22 kg/m² for the active group and 26.9 ± 5.21 kg/m² for the sham group. Most participants were White (85.0% in the active group, 95.0% in the sham group), and over half were married or partnered (65.0% in the active group, 57.9% in the sham group) and possessed education beyond a 2-year college degree (65.0% in the active group, 60.0% in the sham group). No significant baseline differences were noted between the two groups (p > .05). Additionally, baseline comparisons of pain measures revealed no statistically significant differences (p > .05; see Supplemental Table 2). Participants were satisfied with home-based tDCS treatment, expressing appreciation for its enhanced safety, adherence monitoring, low burden, and ease of use.

Table 1 presents fit statistics for unconditional LTA models with two to four statuses. BIC favored a 2-status model as the best solution, while other statistics recommended a 4-status model. The 3-status and 4-status models had one status with a notably small number of participants (e.g., n = 5, 6, respectively) and proved challenging to interpret. Inspection of the 2-status model indicated distinct and meaningful pain patterns. Consequently, we selected the 2-status model as the final configuration for our LTA model.

Table 2 presents the multi-group LTA results. First, we labeled each status based on its item-response probabilities (i.e., the mean NRS and MOBID-2 scores in this study’s context); the first status was labeled “low pain,” while the second was designated “high pain.” As the LTA model was specified to include measurement invariance over time, the definitions of the statuses remained consistent throughout the study. In both groups, the prevalence of “low pain” status generally increased over time, whereas that of “high pain” decreased until 1-month follow-up. Conversely, a slightly opposite trend was observed in the active group from 1 to 3 months’ follow-up.

Table 3 presents the group-specific transition probabilities across the groups. Active group participants with “high-pain” status at baseline had a 52.9% probability of remaining in that status and a 47.1% probability of transitioning to “'low-pain” status immediately after intervention completion. In contrast, no transitions were observed between any statuses in the sham group during this period.

From immediately after intervention completion to the 1-month follow-up, active group participants with “high pain” status had a 55.7% probability of remaining in that status and a 44.3% probability of transitioning to “low pain” status. Furthermore, active group participants with “low pain” status were likely to remain in that status (88.3% probability), while having a 12.1% likelihood of transitioning to “high pain” status. Among sham group participants, those with “high pain” status had a 39.2% likelihood of transitioning to “low pain” status and a 60.8% likelihood of remaining in that status.

From 1 to 3 months’ follow-up, active group participants with “high pain” status had a 58.4% likelihood of transitioning to “low pain” status, while sham group participants essentially exhibited no likelihood of transitioning between statuses. Finally, it is worth noting that certain participants in the active group experienced a regression from "low pain" to "high pain" status. Specifically, those who initially reported “low pain” status immediately after intervention completion had a 12.1% likelihood of regressing to “high pain” status by the 1-month follow-up. Similarly, participants who reported “low pain” status at the 1-month follow-up had a 20.3% likelihood of regressing to “high pain” status by the 3-month follow-up.

The ability of multi-group LTA to allow the simultaneous measurement of both self-reported pain and caregiver-observed pain behaviors in a single model and the exploration of their changes as a function of intervention exposure facilitated a more objective and comprehensive assessment of tDCS’s effects on pain reduction in older adults with ADRD. Overall, active tDCS demonstrated better effects on clinical pain improvement compared with the sham stimulation condition both during and following the intervention. This finding underscores the feasibility and sustainability of our home-based modality, facilitated by caregiver assistance, for pain management in older adults with ADRD. This conclusion aligns with existing literature advocating for home-based tDCS approaches for the current study’s demographic.^37-39

Notably, approximately half of the participants who completed active tDCS transitioned from “high pain” to “low pain” status immediately after intervention completion, whereas those in the sham tDCS group exhibited no pain reduction during this interval. This is particularly important, given the pressure in healthcare to resolve issues as quickly as possible. Yet interestingly, from immediate to the 1-month follow-up, both groups exhibited positive transitions; the sham tDCS group also experienced benefits, albeit with slightly smaller changes (39.2% probability) compared to the active tDCS group (44.3% probability). These results suggest that a sham protocol, previously considered inactive, may potentially exert neuromodulatory effects,⁴⁰ which is consistent with findings from a few prior studies.^41-43 Moreover, we cannot rule out the placebo effect.

Active tDCS participants experienced not only immediate pain reduction post-intervention, but also maintained this improvement for an additional 2 months until the 3-month follow-up period. In addition to the positive transition observed until the 1-month follow-up, approximately 60% of participants in the active tDCS group, who had not shown any improvement by the 1-month follow-up, transitioned to “low pain” status by the 3-month follow-up. In contrast, participants in the sham tDCS group exhibited no positive transitions during this interval. While future studies are warranted to fully elucidate the sustained effectiveness of tDCS over a longer duration, this study’s overall findings represent a promising initial step towards investigating the use of tDCS in a large-scale study aimed at examining its long-term impacts on this population.

Furthermore, the observed overall pain reduction over the 3-month follow-up period in the active tDCS group highlights the importance of determining the optimal dosage of tDCS to achieve sustained effects in older adults with ADRD. Various factors, such as treatment duration and intervals, stimulation parameters, electrode polarity, target brain area, and electrode preparation, can influence the optimal effects of tDCS.⁴⁴ In this study, a daily regimen of a constant 2 mA current for 20 minutes over five consecutive days proved effective both during and after the intervention, particularly suggesting the potential long-term benefits of this specific tDCS approach. However, due to the small sample size used, future research should focus on replicating these findings with larger populations and further investigating the aforementioned optimal factors for maximizing pain relief in the study’s demographic. Simultaneously, investigating the underlying mechanisms of tDCS-induced pain reduction among older adults with ADRD would greatly improve its use in clinical settings.

It is noteworthy that a small proportion of participants in the active tDCS group regressed from “low pain” to “high pain” status post-intervention. This finding underscores the importance of providing timely support to those at higher risk of regression, which can ultimately contribute to the development of sustainable tDCS interventions for older adults with ADRD. However, LTA does not statistically allow for the identification of individuals exhibiting specific transition patterns. Future research based on more advanced experimental designs should focus on understanding the profiles or group characteristics of these individuals (e.g., initial pain intensity or cognitive function that may influence the duration of tDCS effects or overall intervention efficacy) to guide targeted interventions and improve monitoring strategies.

Limitations & future directions

We disclose important limitations. First, owing to small sample size, our findings remain preliminary; statistically, small sample sizes potentially limit stability across time points, possibly causing model identification issues in the cross-sectional component of LTA and affecting the statistical power of the analysis.⁴⁵ Additionally, partly due to the sample size, we selected the 2-status model, which may have overlooked subtle changes. For example, participants in the sham group, where no transition was observed, might have shifted from “high” to “moderate” pain status if the latter had been identified as a category of pain status in this study. Furthermore, we were limited by a relatively homogeneous sample, primarily comprising White and well-educated individuals. Future studies should engage a larger, more diverse demographic to evaluate the intervention. Second, most of the multi-group LTA findings were relatively descriptive; each status’s transition probabilities did not have any verified statistical significance associated with them. Hence, our conclusions must be taken with caution. Lastly, we cannot rule out the possibility that external factors, such as medication use or the introduction of new types of pain treatment, may have influenced the observed sustained effects post-intervention. These factors could potentially confound the effects of tDCS, making it challenging to assert the presence of a sustainable effect decisively. Therefore, caution is warranted before drawing any definitive conclusions.

Finally, a critical missing part of this study is the determination of which participants, based on their characteristics, exhibit greater pain reduction or responsiveness to tDCS. This can be achieved by allowing covariates (time-invariant/time-varying covariates) to interact on the latent status (δ’s) and transition (τ’s) probabilities using an internal model-based approach, that is, LTA with covariates.⁴⁶For instance, by employing this approach, researchers can investigate how race/ethnicity (i.e., time-invariant covariates) or medical conditions such as mental health disorders, including depression (i.e., time-varying covariates), are differentially associated with membership in pain profiles and transition probabilities over time in patients undergoing tDCS. While this approach will be of paramount essence in further tailoring intervention strategies, it was not going to be feasible in the present study as it necessities a relatively large sample size. In cases where the sample size is small, some transitions between statuses may be less frequent from one time point to another, and certain statistics (e.g., odds ratios) cannot be accurately estimated.⁴⁶ Again, future studies with larger sample are encouraged.

Using LTA, this study observed the effects of home-based, remotely supervised tDCS on pain reduction in community-dwelling older adults with ADRD. However, our preliminary conclusions should be interpreted with caution due to the exploratory nature of this study and the small sample size.

ADRD, Alzheimer’s Disease and Related Dementias; AIC, Akaike Information Criterion; BIC, Bayesian Information Criterion; LL, Log-Likelihood; LTA, Latent Transition Analysis; MOBID-2, Mobilization-Observation-Behavior-Intensity-Dementia; NRS, Numeric Rating Scale; SABIC, Sample-Size Adjusted BIC; tDCS, Transcranial Direct Current Stimulation

Consent for publication

Not applicable.

Funding

This research was funded by the NIH/NINR Grant R15NR018050 and R01NR019051.

Acknowledgements

The authors would like to extend their sincere thanks to all participants who took part in the study.

Availability of data and materials

The datasets generated and/or analyzed during the current study are not publicly available (for the protection of participants’ personal information).

Declaration of Generative AI and AI-assisted technologies in the writing process

The authors did not use generative AI tools other than basic tools for checking grammar and spelling.

CRediT authorship contribution statement

Chiyoung Lee: Conceptualization, Writing - original draft, Investigation, Methodology, Formal analysis, Juyoung Park: Conceptualization, Investigation, Supervision, Writing - original draft, Writing - review & editing, Mindy Fain: Investigation, Resources, Validation, Writing - review & editing, James Galvin: Investigation, Resources, Validation, Lindsey Park: Data curation, Investigation, Resources, Validation, Hyochol Ahn: Conceptualization, Funding acquisition, Investigation, Project administration, Resources, Supervision, Writing - review & editing

Lin RJ, Siegler EL. Acute pain management in older adults. In: Cordts GA, Christo PJ, eds. Effective treatments for pain in the older patient. New York: Springer; 2019. pp. 35–52.
Smith PD, et al. Associations among pain, depression, and functional limitation in low-income, home-dwelling older adults: An analysis of baseline data from CAPABLE. Geriatr Nurs. 2016;37(5):348-352.
Smith PD, et al. Communication between office‐based primary care providers and nurses working within patients' homes: an analysis of process data from CAPABLE. J Clin Nurs. 2016;25(3-4):454-462.
Whitlock EL, et al. Association between persistent pain and memory decline and dementia in a longitudinal cohort of elders. JAMA Intern Med. 2017;177(8):1146-1153.
Jakobsson ULF, Hallberg IR. Pain and quality of life among older people with rheumatoid arthritis and/or osteoarthritis: a literature review. J Clin Nurs. 2002;11(4):430-443.
O'Sullivan K, Kennedy N, Purtill H, Hannigan A. Understanding pain among older persons: Part 1—the development of novel pain profiles and their association with disability and quality of life. Age Ageing. 2017;46(1):46-51.
Wang J, et al. Pain management in home health care: relationship with dementia and facility admissions. Pain Manag Nurs. 2021;22(1):36-43.
Grol-Prokopczyk H. Sociodemographic disparities in chronic pain, based on 12-year longitudinal data. Pain. 2017;158(2):313-322.
Smith D, Wilkie R, Croft P, McBeth J. Pain and mortality in older adults: the influence of pain phenotype. Arthritis Care Res. 2018;70(2):236-243.
Boltz M, et al. Pain incidence, treatment, and associated symptoms in hospitalized persons with dementia. Pain Manag Nurs. 2021;22(2):158-163.
Corbett A, et al. Assessment and treatment of pain in people with dementia. Nat Rev Neurol. 2012;8(5):264-274.
Wang J, et al. Pain interference and depressive symptoms in communicative people with Alzheimer's disease: a pilot study. Aging Ment Health. 2018;22(6):808-812.
Achterberg W, Lautenbacher S, Husebo B, Erdal A, Herr K. Pain in dementia. Pain Rep. 2020;5(1):803
Harvey MP, et al. Relieving chronic musculoskeletal pain in older adults using transcranial direct current stimulation: Effects on pain intensity, quality, and pain-related outcomes. Front Pain Res. 2022;3:817984.
Kim S, et al. Effects of transcranial direct current stimulation on physical and mental health in older adults with chronic musculoskeletal pain: a randomized controlled trial. Eur Geriatr Med. 2022;13(4):959-966.
Ahn H, et al. Efficacy of combining home-based transcranial direct current stimulation with mindfulness-based meditation for pain in older adults with knee osteoarthritis: A randomized controlled pilot study. J Clin Neurosci. 2019;70:140-145.
Suchting R, et al. Changes in experimental pain sensitivity from using home-based remotely supervised transcranial direct current stimulation in older adults with knee osteoarthritis. Pain Med. 2020;21(11):2676-2683.
David MCM, Moraes AAD, Costa MLD, Franco CIF. Transcranial direct current stimulation in the modulation of neuropathic pain: A systematic review. Neurol Res. 2018;40(7):557-565.
Knotkova H, Clark WC, Mokrejs P, Padour F, Kuhl J. What do ratings on unidimensional pain and emotion scales really mean? A Multidimensional Affect and Pain Survey (MAPS) analysis of cancer patient responses. J Pain Symptom Manage. 2004;28(1):19-27.
Collins LM, Lanza ST. Latent class and latent transition analysis: With applications in the social, behavioral, and health sciences. Vol. 718. John Wiley & Sons; 2009.
Velicer WF, Martin RA, Collins LM. Latent transition analysis for longitudinal data. Addiction. 1996;91(12s1):197-210.
Wang J, et al. Pain and the Alzheimer's disease and related dementia spectrum in community-dwelling older Americans: A nationally representative study. J Pain Symptom Manage. 2022;63(5):654-664.
Ahn H, et al. Efficacy of transcranial direct current stimulation over primary motor cortex (anode) and contralateral supraorbital area (cathode) on clinical pain severity and mobility performance in persons with knee osteoarthritis: An experimenter-and participant-blinded, randomized, sham-controlled pilot clinical study. Brain Stimul. 2017;10(5):902-909.
Martorella G, Mathis K, Miao H, Wang D, Park L, Ahn H. Self-administered transcranial direct current stimulation for pain in older adults with knee osteoarthritis: a randomized controlled study. Brain Stimul. 2022;15(4):902-909.
Im JJ, et al. Effects of 6-month at-home transcranial direct current stimulation on cognition and cerebral glucose metabolism in Alzheimer's disease. Brain Stimul. 2019;12(5):1222-1228.
Lefaucheur JP, et al. Evidence-based guidelines on the therapeutic use of transcranial direct current stimulation (tDCS). Clin Neurophysiol. 2017;128(1):56-92.
Knotkova H, et al. Automatic M1-SO montage headgear for transcranial direct current stimulation (TDCS) suitable for home and high-throughput in-clinic applications. Neuromodulation. 2019;22(8):904-910.
Fregni F, et al. A randomized, sham‐controlled, proof of principle study of transcranial direct current stimulation for the treatment of pain in fibromyalgia. Arthritis Rheum. 2006;54(12):3988-3998.
Gandiga PC, Hummel FC, Cohen LG. Transcranial DC stimulation (tDCS): a tool for double-blind sham-controlled clinical studies in brain stimulation. Clin Neurophysiol. 2006;117(4):845-850.
Lichtner V, et al. Pain assessment for people with dementia: A systematic review of systematic reviews of pain assessment tools. BMC Geriatr. 2014;14:1-19.
Husebo BS, Ostelo R, Strand LI. The MOBID-2 pain scale: Reliability and responsiveness to pain in patients with dementia. Eur J Pain. 2014;18(10):1419–1430.
Husebo BS, et al. Mobilization-Observation-Behavior-Intensity Dementia Pain Scale (MOBID): Development and validation of a nurse-administered pain assessment tool for use in dementia. J Pain Symptom Manage. 2007;34(1):67–80.
Nylund-Gibson K, et al. Ten frequently asked questions about latent transition analysis. Psychol Methods. 2023;28(2):284.
Lanza ST, Bray BC. Transitions in drug use among high-risk women: an application of latent class and latent transition analysis. Adv Appl Stat Sci. 2010;3(2):203-235.
Nylund-Gibson K, Grimm R, Quirk M, Furlong M. A latent transition mixture model using the three-step specification. Struct Equ Modeling. 2014;21(3):439-454.
Enders CK, Bandalos DL. The relative performance of full information maximum likelihood estimation for missing data in structural equation models. Struct Equ Modeling. 2001;8(3):430-457.
Kim J, Park S, Kim H, Roh D, Kim DH. Home-based, Remotely Supervised, 6-Week tDCS in Patients With Both MCI and Depression: A Randomized Double-Blind Placebo-Controlled Trial. Clin EEG Neurosci. 2023.
Park J, et al. Effect of home-based transcranial direct current stimulation on cognitive function in patients with mild cognitive impairment: a two-week intervention. Yonsei Med J. 2024;65(6):341-347.
Park J, Oh Y, Chung K, Kim KJ, Kim CO, Park JY. Effect of home-based transcranial direct current stimulation (tDCS) on cognitive function in patients with mild cognitive impairment: a study protocol for a randomized, double-blind, cross-over study. Trials. 2019;20:1-9.
Fonteneau C, et al. Sham tDCS: A hidden source of variability? Reflections for further blinded, controlled trials. Brain Stimul. 2019;12(3):668-673.
Braga M, et al. The role of expectation and beliefs on the effects of non-invasive brain stimulation. Brain Sci. 2021;11(11):1526.
Creutzfeldt OD, Fromm GH, Kapp H. Influence of transcortical dc currents on cortical neuronal activity. Exp Neurol. 1962;5(6):436-452.
Nikolin S, Martin D, Loo CK, Boonstra TW. Effects of TDCS dosage on working memory in healthy participants. Brain Stimul. 2018;11(3):518-527.
Woods AJ, et al. A technical guide to tDCS, and related non-invasive brain stimulation tools. Clin Neurophysiol. 2016;127(2):1031-1048.
Pat-Horenczyk R, et al. Stability and transitions in posttraumatic growth trajectories among cancer patients: LCA and LTA analyses. Psychol Trauma. 2016;8(5):541.
Lanza ST, Dziak JJ, Huang L, Wagner A, Collins LM. Proc LCA & Proc LTA users’ guide (version 1.3. 2). University Park: The Methodology Center, Penn State. 2015.

Table 1 Fit statistics for unconditional LTA models.

No. of statuses	AIC	BIC	SABIC	LL	Entropy
2	2043.93	2071.30	2103.39	-1016.65	0.88
3	2032.15	2089.58	1983.18	-982.08	0.89
4	2020.89	2113.78	1941.67	-955.45	0.94

Note. AIC, Akaike Information Criteria; BIC, Bayesian Information Criteria; SABIC, Sample-size Adjusted BIC; LL, Log-Likelihood.

Optimal values in each column are written in bold font.

Table 2 The results of the unconditional LTA model with three latent statuses.

	Latent status
	“Low pain”	“High pain”
Symptom indicator means^a
Self-reported pain (NRS)	21.4 ± 2.48	65.9 ± 3.10
Caregiver-observed pain behaviors (MOBID-2)	1.7 ± 0.21	5.2 ± 0.36
Latent status membership prevalence
Active tDCS
Time 1 (baseline)	0.513	0.487
Time 2 (immediately after intervention completion)	0.742	0.258
Time 3 (1-month follow-up)	0.767	0.233
Time 4 (3-month follow-up)	0.747	0.253
Sham tDCS
Time 1 (baseline)	0.749	0.251
Time 2 (immediately after intervention completion)	0.749	0.251
Time 3 (1-month follow-up)	0.797	0.203
Time 4 (3-month follow-up)	0.797	0.203

^aSymptom indicator means constrained to be equal at baseline, immediately after intervention completion, and at one 1-month, and three 3-months month follow-ups.

Abbreviation. NRS = numeric rating scale; MOBID = mobilization-observation-behavior-intensity-dementia; tDCS = transcranial direct current stimulation

Transition probabilities in bold font correspond to membership in the same latent status at both times.

Table 3 Group-specific transition probabilities across the study groups.

	Latent status
	“Low pain”	“High pain”
Transition probabilities^a (rows for baseline, columns for immediately after intervention completion)
Active tDCS
	“Low pain”	“High pain”
“Low pain”	1.000	0.000
“High pain”	0.471	0.529
Sham tDCS
	“Low pain”	“High pain”
“Low pain”	1.000	0.000
“High pain”	0.000	1.000
Transition probabilities^a (rows for immediately after intervention completion, columns for 1-month follow-up)
Active tDCS
	“Low pain”	“High pain”
“Low pain”	0.879	0.121
“High pain”	0.443	0.557
Sham tDCS
	“Low pain”	“High pain”
“Low pain”	0.933	0.067
“High pain”	0.392	0.608
Transition probabilities^a (rows for 1-month follow-up, columns for 3-month follow-up)
Active tDCS
	“Low pain”	“High pain”
“Low pain”	0.797	0.203
“High pain”	0.584	0.416
Sham tDCS
	“Low pain”	“High pain”
“Low pain”	1.000	0.000
“High pain”	0.000	1.000

^aTransition probabilities in bold font correspond to membership in the same latent status at both times.

Abbreviation. tDCS = transcranial direct current stimulation

The authors declare no competing interests.

SupplementalmaterialsLTAGN.docx

Immediate and sustainable effects of transcranial direct current stimulation on pain reduction in older adults with Alzheimer’s disease and related dementias: a pilot study

Status:

Version 1

Abstract

Figures

Introduction

Study aim

Methods

Design

Participants

Intervention

Measurement

Statistical analysis

Results

Discussion

Limitations & future directions

Conclusion

Abbreviations

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1