An interventional study of Language activities active inference and Language activities passive in Promoting Active Inference of Healthy Older Adults With Language Activity, sponsored by National Taiwan University Hospital. Recruiting at 1 site in Taiwan. Open to participants aged 60 Years and older, including healthy volunteers. Per ClinicalTrials.gov, last updated 2026-01-07.
Sponsored by National Taiwan University Hospital · Not applicable, Interventional, and Treatment
Taiwan is fast approaching a super-aged society, making it urgent to bolster cognitive health in otherwise healthy older adults. This integrated project tackles that need with a language-centered intervention grounded in predictive-coding and active-inference theory. Over 12 weeks, community-dwelling adults aged 65 + join small-group reading-and-writing workshops that train them to actively predict, monitor, and revise linguistic information. Ninety volunteers are randomly allocated to an active language-prediction group, a passive reading group, or a hobby board-game control. Before and after the course, researchers collect behavioural tests, EEG, fMRI, and AI-based speech-language analytics to quantify gains and transfer effects across cognition, emotion, and daily function.
Taiwan will enter a super-aged society in 2025, with adults aged 65 and above expected to comprise 20 % of the population-a shift that makes preserving cognitive health an urgent public-health priority. To meet this challenge, our integrated project draws on predictive-coding and active-inference theory to couple language science with neuroscience, psychology, occupational therapy, and large-language-model analytics. Within this framework, the present sub-project tests a 12-week, small-group language-training curriculum that meets once a week for two hours. Sessions weave together reading aloud, listening, oral summarising, writing, and guided discussion, all structured to elicit continual prediction, self-monitoring, feedback, and creative revision during language use.
Approximately ninety community-dwelling adults aged 65 years or older are being recruited and randomly assigned, in equal blocks of thirty, to one of three arms: an active language-prediction group that emphasises hypothesis-testing during language production and comprehension; a passive reading group that provides leisure reading and sharing without explicit predictive feedback; and a board-game control group that engages cognition but not language prediction.
Before and after the intervention, each participant completes a comprehensive battery. Behaviourally, the investigators assess verbal fluency, language memory, and real-time sentence-prediction accuracy, while electrophysiology (N400, anterior-positivity ERPs, anterior P2, and P300) and functional MRI capture neural plasticity associated with sentence processing, visual reasoning, and cognitive flexibility. Executive functions, divergent and convergent thinking, attention, affect, social cognition, and instrumental activities of daily living provide secondary endpoints to gauge transfer beyond language. Throughout the 12 weeks, spoken narratives are recorded and analyzed for idea density and syntactic complexity with large-language-model tools, yielding fine-grained markers of change unavailable to traditional testing.
The investigators anticipate that targeted, prediction-based training will rejuvenate age-diminished language-prediction signals, drive adaptive reorganization of executive and language networks, and produce broader cognitive and functional gains than either passive reading or non-linguistic gaming. Embedded in community settings, this program also seeks to foster lasting reading habits, strengthen collective cognitive resilience, and offer an ecologically valid blueprint for scaling evidence-based cognitive-health interventions.
243 studies on the registry are indexed under Language; 76 are open to participants now.
This study's planned enrollment of 120 is above the median of 65 across 158 interventional studies indexed under Language.
Browse Language studies →National Taiwan University Hospital is the lead sponsor of 2,615 studies on the registry; 576 are open to participants now.
Of its 11 completed or terminated interventional studies of FDA-regulated products, 2 (18%) have results posted.
Counted across the registry records on this site, refreshed daily.
Go through the criteria one by one — yes, no or not sure — and take the list to your doctor or the study team. Kept in this browser only.
Exclusion Criteria:
Underlined terms in the criteria open these too. Explained by clinicaltrials.gg, not part of the registry record. The study team decides who can take part.
Short tests of memory and thinking, each scored out of 30. Lower scores suggest more difficulty; criteria give a range, either to include people with memory problems or to leave them out.
A reason a treatment, test or procedure should not be given to a particular person, such as an allergy, another illness or a medicine they take.
MMSE and MoCA
Short tests of memory and thinking, each scored out of 30. Lower scores suggest more difficulty; criteria give a range, either to include people with memory problems or to leave them out.
Explained by clinicaltrials.gg, not part of the registry record. The study team decides who can take part.
Contraindication
A reason a treatment, test or procedure should not be given to a particular person, such as an allergy, another illness or a medicine they take.
Explained by clinicaltrials.gg, not part of the registry record. The study team decides who can take part.
Participants will undergo language activities and will be encouraged to actively infer based on the given language materials.
Behavioral: Language activities active inference
Participants will undergo language activities and will receive step-by-step guidance.
Behavioral: Language activities passive
Participants will play boards games under a schedule matching the Experimental and Active Comparator arms.
Behavioral: Board Games
Participants will join reading or writing activities and come up with their own inference about wording or plots.
Participants will join reading or writing activities and will be given step-by-step explanations about the wording and plots.
Participants will play board games with each other.
Changes of neural functional activity during inferential processing
Participants will undergo a Rule Inference fMRI task to infer underlying rules that map color configurations of circles in a triangular arrangement to a target color category within as few tries as possible under active or passive conditions. The goal for participants will be to infer the cue-category association rules using as few cues as possible. The primary outcome measure here is the degree of neural response estimate change in blood oxygen level dependent (BOLD) signal pre- and post-intervention.
Time frame: Week 0, Week 12
Changes of overall accuracy during inferential processing
Changes from pre- to post-intervention in participant overall accuracy in identifying latent rules in the Rule Inference fMRI task.
Time frame: Week 0, Week 12
Changes of learning rate during inferential processing
Changes from pre- to post-intervention in participant number of trials to criterion in the Rule Inference fMRI task.
Time frame: Week 0, Week 12
Changes of strategic performance during inferential processing
Changes from pre- to post-intervention in participant coefficients of expression of modeled response strategies in the Rule Inference fMRI task will be assessed.
Time frame: Week 0, Week 12
Changes of functional electrical brain responses during reading
Participants will undergo EEG language reading tasks in which participants were told to actively comprehend the sentence and to infer meanings for unknown pseudowords. The primary outcome measure here is the changes in N400 responses and anterior positive responses pre- and post-intervention.
Time frame: Week 0, Week 12
Changes in the Montreal Cognitive Assessment (MoCA) score
Pre- to post-intervention changes in participant MoCA score. Score range from 0 to 30 with higher scores indicating better cognitive ability.
Time frame: Week 0, Week 12
Change in Wechsler Memory Scale III Logical Memory I & II
Score range 0 - 75. Higher score indicates better verbal episodic memory.
Time frame: Week 0, Week 12
Change in Wechsler Memory Scale III Face Memory
Score range 0 - 48. Higher score indicates better visual face memory.
Time frame: Week 0, Week 12
Change in Wechsler Memory Scale III Verbal Paired Memory
Score range 0 - 32. Higher score indicates better verbal memory and learning.
Time frame: Week 0, Week 12
Change in Wechsler Memory Scale III Family Pictures I & II
Score range 0 - 64. Higher score indicates better visual memory and learning.
Time frame: Week 0, Week 12
Change in Wechsler Memory Scale III Word Lists I & II
Score range 0 - 36. Higher score indicates better verbal memory and learning. For II, recall score range is 0 to 8; recognition score range is 0 to 24.
Time frame: Week 0, Week 12
Change in Wechsler Memory Scale III Visual Reproduction I & II
Score range 0 - 104. Higher score indicates better visual memory. For II, recall score range is 0-104; recognition score range is 0-48.
Time frame: Week 0, Week 12
Change in Wechsler Memory Scale III Spatial Span
Score range 0 - 32. Higher score indicates better spatial memory.
Time frame: Week 0, Week 12
Change in Wechsler Memory Scale III Digit Span
Score range 0 - 32. Higher score indicates better auditory memory.
Time frame: Week 0, Week 12
Change in Wechsler Adult Intelligence Scale III Vocabulary
Score range 0 - 66. Higher score indicates better vocabulary.
Time frame: Week 0, Week 12
Change in Wechsler Adult Intelligence Scale III Digit Symbol
Score range 0 - 133. Higher score indicates better processing speed.
Time frame: Week 0, Week 12
Change in Wechsler Adult Intelligence Scale III Block Design
Score range 0 - 68. Higher score indicates better visual processing.
Time frame: Week 0, Week 12
Change in Wechsler Adult Intelligence Scale III Arithmetic
Score range 0 - 22. Higher score indicates better mathematical computation ability.
Time frame: Week 0, Week 12
Change in Wechsler Adult Intelligence Scale III Matrix Reasoning
Score range 0 - 26. Higher score indicates better reasoning.
Time frame: Week 0, Week 12
Changes of overall accuracy for meaning inferences
Changes from pre- to post-intervention in participants' overall accuracy in judging the semantic relation between newly learned pseudo-words and their semantic associates.
Time frame: Week 0, Week 12
Changes of electrical brain responses during the task switching experiment
Participants will undergo an EEG task-switching paradigm in which they are instructed to make an odd/even judgment or a size judgment on presented numbers.
Time frame: Week 0, Week 12
Changes of overall accuracy during the task-switching paradigm
Changes from pre- to post-intervention in participants' switch cost, assessed through overall accuracy.
Time frame: Week 0, Week 12
Changes of Originality in Alternate Use Test (AUT)
Developed by Guilford et al. (1967), this tool serves as a quantitative measure for assessing creativity and divergent thinking. In under three minutes, participants are given common multipurpose items and are asked to creatively think of uses for these items on paper, such as various uses for a brick. The scoring of Originality measures the uniqueness of the response in comparison to others. For instance, an answer that only 1% of participants come up with scores 2 points; an answer given by 5% scores 1 point.
Time frame: Week 0, Week 12
Changes in Chinese Remote Association Task (CRAT)
The Chinese Remote Association Task is used to measure individuals' abilities in convergent thinking. The original developer of the Remote Associates Test (RAT) was Mednick (1962). The Chinese version of the Remote Associates Test was developed by Huang et al. (2012). In this test, participants are provided with three Chinese words (e.g., Taiwan, land, barrier), and they are required to find a fourth Chinese word that forms a meaningful association with all three given words (e.g., strait). The study consists 30 questions and includes two versions of the test, A and B, which have a correlation of 0.610. The internal consistency coefficients for these versions are 0.808 and 0.801, respectively （Ren et al., 2004).
Time frame: Week 0, Week 12
Changes of completion time in Color Trails Test (CTT)
Performance is measured by recording the completion time.
Time frame: Week 0, Week 12
Changes of number of trials administered in Wisconsin Card Sorting Test (WCST)
Participants' performance is evaluated by measuring total trials required to complete the task (fewer trials indicate better performance).
Time frame: Week 0, Week 12
Changes of overall fluency of language production
Changes from pre- to post-intervention in participants' fluency in language production.
Time frame: Week 0, Week 12
Changes of the number of errors in Color Trails Test (CTT)
The Color Trails Test (CTT) is a neuropsychological assessment designed to measure sustained attention, alternating attention, and executive functioning (D'Elia et al., 1996). The test typically takes between 3 and 8 minutes to complete and consists of two parts. In Part 1, participants are instructed to connect numbered circles (1 through 25) in ascending order as quickly as possible. In Part 2, participants must again connect the numbered circles in order but are required to alternate between pink and yellow circles, adding a set-shifting and divided attention component. Performance is measured by the number of errors for each part. This study employed the Chinese adaptation of the CTT developed by Guo (2009), which includes normative data for Chinese adults aged 50 to 88 years. The test-retest reliability coefficients for Part 1 and Part 2 were reported as 0.618 and 0.833, respectively.
Time frame: Week 0, Week 12
Changes of Fluency in Alternate Use Test (AUT)
Developed by Guilford et al. (1967), this tool serves as a quantitative measure for assessing creativity and divergent thinking. In under three minutes, participants are given common multipurpose items and are asked to creatively think of uses for these items on paper, such as various uses for a brick. The scoring of Fluency measures the total number of responses.
Time frame: Week 0, Week 12
Changes of Flexibility in Alternate Use Test (AUT)
Developed by Guilford et al. (1967), this tool serves as a quantitative measure for assessing creativity and divergent thinking. In under three minutes, participants are given common multipurpose items and are asked to creatively think of uses for these items on paper, such as various uses for a brick. The scoring of flexibility is as follow: Responses are categorized based on their characteristics, and the number of categories is counted as the score.
Time frame: Week 0, Week 12
Changes of overall reaction time during the task-switching paradigm
Changes from pre- to post-intervention in participants' switch cost, assessed through overall reaction time.
Time frame: Week 0, Week 12
Changes of percent correct in Wisconsin Card Sorting Test (WCST)
The proportion of trials with correct responses is measured.
Time frame: Week 0, Week 12
Changes of percent errors in Wisconsin Card Sorting Test (WCST)
The proportion of trials with incorrect responses is measured.
Time frame: Week 0, Week 12
Changes of number of categories completed in Wisconsin Card Sorting Test (WCST)
The number of sorting categories completed is measured (i.e., 10 consecutive correct responses per rule).
Time frame: Week 0, Week 12
Changes of trials to complete first category in Wisconsin Card Sorting Test (WCST)
The number of trials required to achieve the first sorting criterion is measured.
Time frame: Week 0, Week 12
Changes of percent perseverative responses in Wisconsin Card Sorting Test (WCST)
The percent perseverative responses that continued to follow the previous rule, regardless of correctness, is measured.
Time frame: Week 0, Week 12
Changes of percent perseverative errors in Wisconsin Card Sorting Test (WCST)
The proportion of incorrect responses due to perseveration (i.e., failure to shift after rule change) is measured.
Time frame: Week 0, Week 12
Changes of percent nonperseverative errors in Wisconsin Card Sorting Test (WCST)
The proportion of incorrect responses not attributable to perseveration is measured.
Time frame: Week 0, Week 12
Plan to share: Yes — Data that can be shared include anonymized neuropsychological assessment scores, cognitive behavioral performance scores, neurophysiological data that have been published.
Supporting information: Study protocol, Statistical analysis plan, Analytic code
Time frame: Data will become available 1 year after primary results are published by the central research team. Data are anticipated to be available for sharing for an indefinite period after the above criteria is met.
Access criteria: Data sharing will be done based direct requests and on case-by-case evaluation for appropriateness. Use of shared data will require agreement on appropriate citation of data sources at least or authorship inclusion or acknowledgement.
No publications or documents are linked to this record.
Started Sep 2025 (actual). Primary completion is estimated for Dec 2026, 15 months after the start.
Of 89 non-industry interventional studies with no phase under Language, half reached primary completion within 32 months of starting (the median), counting those still running; a quarter within 17 and three quarters within 64. This study's planned 15 months is below that.
Counting every sponsor, half of 93 studies reached it within 33 months; a quarter within 17 and three quarters within 64.
How these figures are worked out, and which studies are left out of them.
Eligibility is decided by the study team. Share this record with your doctor or contact the team directly.
Contact study teamGet an email when the registry record changes — status, dates, results — or when someone posts here.
Sign in to followQuestions and observations about this study, from anyone following it. Not medical advice, and not a channel to the study team — their contact details are on the registry record.
Sign in to join the discussion. Reading takes no account; posting does. You choose a display name, and a pseudonym is the default.
Nothing here yet. If you are running this trial, taking part in it, or weighing whether to, this is the place to say so.
National Taiwan University Hospital →