How Art Influences Mood and Physiological States

Executive summary

Across disciplines, the best-supported conclusion is that art can change how adults feel in the moment and, under some conditions, how their bodies regulate stress. The strongest evidence is for music-based interventions and structured group arts programs, which show small-to-moderate physiological effects and moderate psychological benefits for stress, depression, and anxiety. Visual art viewing and visual art therapy also show meaningful promise, especially for self-reported stress, anxiety, depressive symptoms, and eudaimonic well-being, but the physiological evidence is thinner and methodologically more uneven than the mood evidence. Theater/live performance and digital/interactive art are scientifically interesting and increasingly studied, but their health-oriented biomarker literature is still comparatively early-stage. Unless otherwise noted, this report assumes adult populations and does not assume any specific setting beyond clinical, community, museum, lab, and daily-life contexts. citeturn41search1turn41search0turn28search0turn36search3turn37search0turn42search7

Physiologically, art-related effects are not limited to a single system. Studies have measured autonomic markers such as heart rate and heart-rate variability, endocrine markers such as cortisol, electrodermal activity and skin temperature, neural activity via EEG, fNIRS, and fMRI, and—more rarely—immune or inflammatory markers such as IL-6, TNF-α, CRP, and immunoglobulin responses. The overall pattern is not “art always calms the body.” A more accurate synthesis is that art can reduce stress biology while also increasing engaged, positively toned arousal. That is why some studies report lower cortisol together with higher skin conductance or more dynamic heart activity during compelling aesthetic experiences. citeturn21search1turn39search0turn41search1turn13search2turn42search8turn17search2

Methodologically, the field has matured but remains uneven. Lab studies offer tight control and clearer causal inference, while ambulatory and ecological momentary assessment studies better capture everyday arts engagement and reduce recall bias. Longitudinal cohort studies suggest that repeated cultural engagement is associated with lower depression risk and greater flourishing, but selection effects and residual confounding remain serious concerns. Randomized controlled trials exist across music therapy, dance movement therapy, museum-based programming, and visual art therapy, yet many are still small, heterogeneous, or underpowered for biomarker endpoints. citeturn32search13turn32search0turn24search4turn22search2turn22search1turn28search0turn34view0

For practice, the evidence supports using arts-based approaches as adjuncts rather than replacements for established care. Official guidance is more supportive where risk is low and personalization is central: NICE advises tailoring well-being-promoting activities for people with dementia and explicitly lists art and music therapy among acceptable options, while NICE’s 2023 stroke rehabilitation guideline concluded that evidence for music therapy after stroke was too limited for a clinical recommendation and instead issued a formal research recommendation. WHO’s social prescribing toolkit provides a systems framework for connecting people to community activities, including arts-based options. citeturn35view0turn35view1turn34view0turn34view1turn33search2

Definitions and scope

In affective science, emotion is usually treated as a more acute, stimulus-linked, and often higher-intensity response, whereas mood is broader, more diffuse, and more sustained over time. This distinction matters because art studies frequently mix outcomes that sit at different timescales: immediate state anxiety, arousal, pleasure, and awe are not identical to longer-duration mood, depressive symptom burden, or flourishing. A rigorous reading of the literature therefore asks not only whether art “helps,” but what kind of feeling state it changes, how quickly, and for how long. citeturn20search0turn20search14turn20search6

For this report, art includes both receptive forms—viewing visual art, listening to music, attending dance or theater, engaging with digital or interactive artworks—and active/creative forms such as drawing, painting, sculpting, singing, instrument playing, dancing, acting, and VR-based creative making. This is broader than psychotherapy-only definitions and aligns better with contemporary arts-and-health research, where mood and physiology are studied not just in therapy rooms but also in museums, concert halls, care homes, online platforms, and real-world daily life. citeturn12search14turn41search1turn29search1turn15search20turn17search6turn28search0

The field’s central analytic problem is that “art” is not a single exposure. The effects of original artworks versus reproductions, solitary versus group participation, active making versus passive viewing, guided versus unguided interpretation, preferred versus prescribed music, and live versus recorded performance can differ materially. Many headline claims about “art reducing stress” are therefore too generic unless they specify modality, context, dose, and target outcome. citeturn42search7turn27search2turn23search0turn41search1turn15search1

The timeline below summarizes selected milestones that shaped current thinking, from neuroaesthetics and psychophysiology to pragmatic interventions and ecological methods. citeturn7search10turn39search0turn22search2turn29search6turn41search1turn27search2turn28search0turn32search0turn42search2

timeline
    title Selected developments in arts, mood, and physiology research
    2009 : Neuroaesthetics review consolidates reward, emotion, and sensorimotor approaches
    2016 : Short visual art-making session shown to lower salivary cortisol in healthy adults
    2018 : Cultural engagement linked to lower later depression risk and better well-being in cohort work
         : Art-making HRV studies begin testing material-specific autonomic effects
    2019 : Meta-analytic support grows for dance movement therapy in adult depression
    2020 : Music meta-analyses quantify physiological and psychological stress effects
         : Multicenter RCT of dance movement therapy for depression published
    2021 : Scoping review concludes visual art viewing has promising but methodologically mixed stress evidence
    2024 : JAMA meta-analysis of active visual art therapy published
         : Portable fNIRS study links museum-related prefrontal engagement with mood and stress change
    2025 : Systematic review of art viewing and well-being synthesizes 38 studies
         : Group arts meta-analysis shows moderate reductions in depression and anxiety in older adults
         : Live performance studies show stronger physiological arousal than identical screen versions
    2026 : Large-scale museum viewing study reports better subjective well-being and lower stress
         : EMA review maps 39 arts, nature, and culture studies in daily life

Physiological markers and study designs

The literature uses a layered measurement strategy. Self-report remains indispensable because mood, meaning, awe, and pleasure are partly subjective. But the more rigorous studies pair self-report with physiological markers so that “feeling better” is not inferred from questionnaires alone. What makes the field difficult is that these markers do not all index the same thing: some capture stress load, some capture arousal, some capture attention, and some capture longer-latency inflammatory responses. citeturn21search1turn21search20turn27search2

Marker	What it most plausibly indexes	Typical collection methods in art studies	What recent evidence tends to show	Main caveats	Representative sources
Heart rate and HRV	Acute autonomic state and vagal flexibility	ECG, chest straps, PPG wearables, ambulatory sensors	Music and some museum/group-art interventions often show lower heart rate or more adaptive variability, but compelling live or original-art experiences can also produce higher, more dynamic arousal signatures without implying distress	Strongly affected by respiration, posture, movement, speech, and physical exertion; HRV needs careful preprocessing	HRV and stress-methods review; music meta-analysis; museum and gallery studies. citeturn21search1turn41search1turn23search1turn42search8
Cortisol and salivary alpha-amylase	HPA-axis stress response and sympathetic activation	Saliva samples before/after exposure, repeated same-day sampling, sometimes multi-day ambulatory collection	Acute reductions are reported after art making, music, and some museum-viewing experiences; effects are plausible but not universal	High diurnal variability; food, caffeine, sleep, and sampling time are major confounders; cortisol and perceived stress do not always covary strongly	Art-making cortisol study; music stress meta-analysis; biomarker review. citeturn39search0turn41search1turn21search21turn21search1
Skin conductance and skin temperature	Sympathetic arousal and orienting	Finger/palm EDA sensors, wearables, event-linked recordings	Often rises during engaging, surprising, or live aesthetic experiences, including when people report more enjoyment or reduced distress overall	High arousal can be pleasant or unpleasant; interpretation requires context and self-report	Live performance study; artwork-label psychophysiology; original-art gallery study. citeturn15search1turn12search17turn42search8
EEG patterns	Attention, approach/avoidance tendencies, oscillatory affective processing	Portable EEG or lab EEG; frontal alpha asymmetry, theta/beta ratio, spectral power	Museum and performance work suggests EEG indices can track attention and anxiety-related change; but frontal alpha asymmetry remains a mixed biomarker rather than a settled one	Motion artifacts, weak specificity, and poor cross-study standardization remain major issues	EEG review in visual neuroaesthetics; museum EEG pilot; FAA reviews/meta-analysis. citeturn31search8turn31search16turn21search10turn21search2
fNIRS and fMRI	Cortical hemodynamics and network-level processing	Portable fNIRS in museums or movement-tolerant settings; scanner-based fMRI in labs	Visual art and music engage sensory systems plus reward and self-referential/default-mode networks; portable fNIRS can capture prefrontal engagement during real museum visits	fNIRS is surface-limited; fMRI is expensive and ecologically constrained	Museum fNIRS study; neuroaesthetics review; fMRI meta-analyses. citeturn17search11turn31search4turn30search0turn7search18
Immune and inflammatory markers	Short-term inflammatory change and immune response	Salivary or blood IL-6, TNF-α, CRP, IgA, related cytokines	Evidence is promising but sparse. Some gallery-viewing and perioperative music studies report reduced inflammatory markers or cortisol changes, but the literature is too thin for firm generalization	Assays are costly; kinetics differ from acute autonomic measures; many medical confounders	Original-art gallery study; music inflammation review; art-therapy biomarker scoping review. citeturn13search2turn19search0turn17search2

Methodologically, the field now spans at least five major designs. Lab experiments maximize control over stimulus, timing, and baseline. Ambulatory studies use wearables and smartphones to track physiology in natural settings. Longitudinal cohort studies estimate whether repeated arts engagement predicts later mood or flourishing. Randomized controlled trials test intervention effects in clinical or community populations. Ecological momentary assessment captures mood and behavior in real time, reducing retrospective bias and making it possible to study within-person dynamics rather than only between-person averages. The 2026 systematic review of EMA studies on arts, nature, and culture identified 39 articles and concluded that these methods are both feasible and analytically valuable for capturing behavior–emotion interactions in daily life. citeturn32search0turn32search7turn32search13

A particularly useful example of daily-life design is the ecological momentary music intervention literature. In a 2022 feasibility study, 10 chronically stressed women were prompted over 18 days to use self-selected music in stressful moments, with repeated self-report and saliva sampling to test both subjective and biological stress change in real world conditions. That design is small, but it exemplifies the direction the field needs: repeated, in-context measures rather than single pre/post lab snapshots. Similarly, experience-sampling work on participatory art collected 2,495 momentary reports from 41 artists, demonstrating that intensive repeated assessment is practical in arts research. citeturn24search4turn24search19turn32search2turn32search9

The main measurement lesson is that single biomarkers are weak judges. Biomarker reviews now emphasize repeated sampling, standardized timing, multimodal batteries, and careful handling of confounds. This matters because perceived stress and cortisol do not always align longitudinally, and because markers such as skin conductance may rise during emotionally rich, positive encounters rather than merely during threat. The best studies therefore use convergent designs: self-report, physiology, context, and ideally behavior or function. citeturn21search1turn21search21turn12search17turn15search1

Mechanisms and theoretical models

The most useful current models are multi-pathway models. Neuroaesthetic accounts no longer treat aesthetic experience as a single “beauty center.” Instead, they describe a sequence in which sensory analysis, valuation, autobiographical relevance, and emotion regulation interact. Reviews and meta-analyses indicate that aesthetic engagement recruits sensory cortices, reward-related regions, and self-referential/default-mode systems, with personally moving art showing especially strong links to self-related processing. citeturn31search4turn7search18turn30search0turn30search1

flowchart LR
    A[Art exposure or art-making] --> B[Sensory processing and attention]
    B --> C[Appraisal, meaning, and interpretation]
    C --> D[Reward and pleasure systems]
    C --> E[Emotion regulation and stress appraisal]
    A --> F[Embodiment, movement, and sensorimotor engagement]
    A --> G[Social connection, synchrony, shared attention]
    D --> H[Autonomic and endocrine change]
    E --> H
    F --> H
    G --> H
    H --> I[Immediate emotion and state-anxiety change]
    I --> J[Repeated engagement may influence mood, flourishing, and health behavior]

A second family of models centers on reward and pleasure systems. Music is the clearest case. A 2019 pharmacological study showed that manipulating dopamine availability changed the pleasure and motivational value people experienced while listening to music, providing rare causal evidence that music reward is not just metaphorically but neurochemically reward-like. More recent neuroimaging synthesis suggests that music and visual art share some valuation architecture but are not neurally identical; music more strongly depends on auditory prediction and reward coupling, while visual art more consistently engages networks tied to visual evaluation and self-referential appraisal. citeturn30search7turn30search3turn30search0turn30search6

A third family emphasizes embodied cognition. Dance, theater, and many forms of visual art making are not purely cognitive exposures. They involve posture, gesture, interoception, rhythmic entrainment, and sensorimotor simulation. This matters because bodily action can change affect directly, not just symbolically. Dance movement therapy research and live-performance physiology support this claim: movement-based interventions improve depression and anxiety, while audiences at live concerts and theater performances show measurable physiological synchrony, indicating shared embodied dynamics across people rather than isolated individual responses. citeturn29search0turn29search1turn15search7turn15search20

A fourth pathway is stress reduction and restoration. In this framework, art changes appraisal, attention, and physiological load in ways that reduce HPA-axis activation and improve autonomic regulation. But the most sophisticated reading is not “sedation.” Many aesthetic experiences seem to combine reduced physiological stress burden with activated engagement, curiosity, or awe. The original-art gallery studies are especially instructive here: they report lower cortisol and inflammatory markers alongside signs of arousal and involvement. That pattern fits theories of adaptive positive engagement rather than quiet passivity. citeturn13search2turn42search8turn21search1

Finally, many intervention studies are best explained by social and therapeutic mechanisms rather than purely sensory ones. Group arts can reduce depression and anxiety through shared activity, identity, routine, belonging, and nonverbal expression. Arts-on-prescription programs also likely work through a bundle of ingredients: social contact, agency, mastery, meaning, and a low-stigma route into help-seeking. Recent syntheses increasingly argue that these social and self-transformative ingredients are not “extras”; they are often the intervention. citeturn23search0turn37search0turn42search7turn17search4

Empirical evidence since 2015

The post-2015 evidence base is large enough to support differentiated conclusions. It does not support a blanket claim that every art form reliably produces the same physiological profile. Instead, the literature points to a hierarchy of confidence: music has the most mature psychophysiological evidence; structured group arts and dance have increasingly persuasive psychological outcome data; visual art viewing has promising but still methodologically mixed evidence; visual art therapy shows benefits in clinical outcomes but trial quality is often low; theater/live performance and digital/interactive modalities show strong affective and synchrony signals, but fewer health-oriented biomarker studies. citeturn41search1turn41search0turn27search2turn28search0turn37search0

Modality or study	Design	Sample	Markers or outcomes	Main finding	Key limitation	Source
Visual art making in healthy adults	Quasi-experimental pre/post	39	Salivary cortisol	Forty-five minutes of art making produced a statistically significant cortisol reduction; about 75% of participants decreased	No control group; healthy convenience sample	Kaimal et al., 2016. citeturn39search0turn39search6
Art museum lunchtime visit	Single-group pre/post	31	Self-reported stress/arousal; saliva cortisol	Stress and arousal decreased subjectively after the museum visit; cortisol did not change	No control group; brief exposure; lunch-hour sample	Ter-Kazarian, 2020 thesis/article. citeturn25search12turn25search6turn25search2
Figurative vs modern art vs office visit	Experimental museum study	77 young women	Blood pressure; heart rate	Visiting figurative art reduced systolic blood pressure; heart rate did not differ pre/post	Narrow demographic; short-term physiology only	Mastandrea et al., 2019. citeturn26search0turn26search3turn26search5
Museum visit with portable fNIRS	Real-world neuroimaging study	19 older adults	Prefrontal hemodynamics; mood/stress	Prefrontal engagement during artwork analysis predicted the well-being effect of the museum visit	Small sample; correlational mechanism test	Dupuy et al., 2024. citeturn17search11turn14search1turn25search0
Original artworks vs reproductions	Between-group experimental gallery study	50 adults	Cortisol, IL-6, TNF-α, heart dynamics, skin temperature	Viewing original art reportedly reduced cortisol by 22% versus 8% for reproductions, with IL-6 and TNF-α reductions and concurrent autonomic engagement markers	Authenticity and gallery setting were bundled; replication needed	Worrell et al., 2025 study report/manuscript. citeturn13search2turn42search8
Online digital art viewing	Randomized brief online exposure	84	Mood, state anxiety, loneliness, well-being	One to two minutes of online art viewing improved negative mood, state anxiety, loneliness, and well-being	Mostly self-report; very brief exposure; physiology absent	Trupp et al., 2022. citeturn12search14
VR vs 2D art making vs VR control	Randomized comparison	44	Heart rate, skin conductance, salivary alpha-amylase, mood/anxiety	All groups improved mood and anxiety; the VR art-making group showed the greatest heart-rate decrease	Novelty effects possible; student sample	Richesin et al., 2021. citeturn14search0turn18search5
Music interventions and stress	Systematic review + meta-analysis	Noted across included studies	Physiological and psychological stress outcomes	Overall stress reduction: physiological d = .380, psychological d = .545; heart rate effects were larger (d = .456) than blood pressure (d = .343) or hormone levels (d = .349)	Heterogeneous interventions and settings	de Witte et al., 2020. citeturn41search1
Music therapy for stress	Multilevel meta-analysis	47 studies, 76 effect sizes, 2,747 participants	Combined stress-related outcomes	Overall medium-to-large effect on stress-related outcomes: d = .723	Many studies involved music listening by healthcare staff rather than standardized therapist-delivered protocols	de Witte et al., 2022. citeturn41search0turn41search2
Dance movement therapy for depression	Multicenter RCT in Finland	109	Depressive symptoms	DMT + treatment as usual reduced depression more than treatment as usual alone at follow-up	Mostly female sample; mood/self-report focus more than biomarkers	Hyvönen et al., 2020. citeturn29search1turn29search4
Dance/movement and chronic stress	Randomized comparison	40 older adults	Cortisol awakening response	Dance/movement training lowered chronic-stress CAR relative to comparator conditions	Small sample; healthy older adults	Vrinceanu et al., 2019. citeturn29search5turn29search17
Group arts for older adults	Systematic review + meta-analysis	39 studies; depression n = 3,360, anxiety n = 949	Depression and anxiety	Depression d = 0.70 before publication-bias adjustment and d = 0.42 after trim-and-fill; anxiety d = 0.76; stronger depression effects in care homes (d = 1.07) than community (d = 0.51)	Publication bias for depression; heterogeneity remains	Quinn et al., 2025. citeturn38search3turn38search1
Active visual art therapy across conditions	Systematic review + meta-analysis	~4,200 total in review; 50 studies meta-analyzed with 2,766 participants	Multiple health outcomes	Visual art therapy showed benefits for some outcomes, but most studies were low quality and many outcomes were unchanged	Heterogeneity and risk of bias limit certainty	Joschko et al., 2024. citeturn28search0turn28search4
Visual art therapy for adult depressive symptoms	Systematic review + meta-analysis	15 RCTs, 932 participants	Depressive symptoms	Significant reduction in depressive symptoms: SMD = -0.73	Evidence rated low quality	Han et al., 2024. citeturn36search3turn36search10

Several patterns stand out from this comparison. First, acute stress reduction is easier to detect than durable mood change, and durable mood change is easier to detect than durable physiological change. Short-term pre/post cortisol or heart-rate shifts are common targets, while long-term immune or neural outcomes are rare. That asymmetry is one reason art can appear more robust in popular reporting than in strict biomarker synthesis. citeturn39search0turn41search1turn19search0

Second, visual art viewing is promising but not yet methodologically settled. The 2021 scoping review of viewing visual artworks on stress concluded that the evidence was promising, but also emphasized major gaps in standardization and the need for stronger RCTs. The 2025 systematic review of art viewing and well-being, which synthesized 38 studies and 6,805 participants, found the clearest convergent support for eudaimonic well-being, while support for other outcomes such as stress and anxiety was less strong and more context-dependent. This is a critical corrective to oversimplified claims that “looking at art works” full stop. citeturn27search2turn27search7turn42search7turn36search4

Third, active making and therapist-led visual art interventions show more consistent symptom benefits than passive viewing alone, especially for depression and anxiety. But those gains come with an evidence-quality warning. The 2024 JAMA Network Open meta-analysis found genuine benefit signals, yet also judged most supporting studies low quality. That means the prudent conclusion is not skepticism, but qualified optimism with a demand for better trials. citeturn28search0turn28search4turn36search3

Fourth, music remains the best-developed psychophysiological art modality. Across meta-analyses, music interventions reduce stress on both physiological and psychological measures, with particularly notable effects for heart rate. Music therapy, as distinct from generic music exposure, shows medium-to-large pooled effects on stress-related outcomes. That does not mean every playlist is therapeutic, but it does mean the music literature currently offers the clearest quantitative case that an art modality can affect both mood and measurable stress physiology. citeturn41search1turn41search0

Fifth, dance and group arts appear especially strong when the mechanism is likely to be social-plus-embodied rather than purely sensory. Meta-analytic and RCT evidence suggests that dance movement therapy improves depression and related outcomes, and subgroup findings from group-arts research show larger effects in care-home settings than in community settings. That pattern is consistent with a model in which bodily movement, social contact, routine, and creative agency all contribute to benefit. citeturn29search1turn29search6turn38search3

Finally, theater/live performance and digital/interactive art are scientifically rich but clinically younger. Live performance studies show stronger emotional and physiological responses than identical screen versions, and audience synchrony work has demonstrated shared autonomic dynamics in concerts and theater. Online and VR art studies show changes in mood and anxiety and, in some cases, heart rate or salivary alpha-amylase, but clinical-grade biomarker evidence is still sparse. citeturn15search1turn15search20turn15search7turn12search14turn14search17turn14search0

Art-based interventions and practical guidance

When the literature moves from “Does art influence mood and physiology?” to “Should we implement art-based interventions?”, the answer becomes more conditional. The state of the evidence supports implementation where risk is low, tailoring is possible, and arts interventions are integrated as complements to—not substitutes for—standard clinical or psychosocial care. That is the practical position most consistent with both the empirical record and official guidance. citeturn23search0turn28search0turn41search0turn35view0turn34view0

Intervention class	Typical populations/settings	Typical “dose” in the literature	Strongest observed outcomes	Physiological evidence	Current evidence posture or guidance	Sources
Visual art therapy	Anxiety, depression, cancer care, rehabilitation, mixed clinical groups	Single-session to multi-week programs; individual or group	Reduction in depressive symptoms and anxiety in several meta-analyses; some gains in self-esteem and functioning	Acute cortisol and HR/HRV changes shown in small studies; biomarker literature still thin	Promising, but systematic reviews repeatedly judge study quality as mixed to low	citeturn28search0turn36search3turn39search0turn17search2
Music therapy and music listening	Perioperative care, stress management, dementia care, neurorehabilitation, community stress reduction	One-off listening sessions to repeated therapist-led programs	Consistent stress reduction, anxiety reduction, and improved emotional regulation; pooled effects are strongest among art modalities	Best-developed physiological literature: heart rate, blood pressure, cortisol, hormone outcomes	Strongest quantitative support among arts modalities; nevertheless implementation details matter	citeturn41search1turn41search0turn19search0
Dance movement therapy	Adults with depression, older adults, community and therapeutic settings	Repeated sessions over weeks to months	Depression/anxiety reduction, QoL gains, body-related and interpersonal benefits	Some evidence for chronic-stress biomarker change such as cortisol awakening response	Effective enough to justify continued clinical use and larger biomarker-focused trials	citeturn29search0turn29search1turn29search5turn29search6
Museum and gallery programs	Older adults, community participants, social prescribing contexts	Brief visits to 3-month cycles of weekly activities	Subjective stress reduction, well-being gains, some cardiovascular effects	Heart-rate and—in newer studies—cortisol/inflammatory shifts reported	Promising, especially as low-risk community intervention; strongest evidence still emerging	citeturn23search1turn23search4turn25search12turn13search2turn42search8
Arts on prescription	Primary care and community mental health; social isolation; mild-to-moderate psychosocial distress	Commonly 6–12 week referral-based community programs	Significant improvement in well-being in meta-analysis; qualitative benefits in social and psychological domains	Biomarker evidence generally absent	Best viewed as scalable community support; compatible with WHO social prescribing implementation frameworks	citeturn23search0turn23search6turn33search2
Digital and VR art	Students, hospital patients, online users, experimental settings	Very brief exposures to multi-session VR programs	Mood, anxiety, loneliness, and engagement benefits in early studies	Emerging heart-rate, EDA, and salivary-marker work; evidence still preliminary	Useful for access and experimentation, but evidence is not yet as mature as for music or in-person group arts	citeturn12search14turn14search17turn14search0turn13search3

For clinicians

Use art-based interventions as adjunctive tools matched to clear targets. If the target is acute stress down-regulation, music and some forms of visual art making have the strongest physiological backing. If the target is depression or anxiety in older adults, group arts and dance/movement therapies have increasingly strong outcome data. If the target is social isolation, agency, and low-threshold support, arts-on-prescription and museum/community referral models are sensible options. NICE explicitly advises tailoring well-being activities for people with dementia to preferences and lists art and music therapy among acceptable examples. citeturn41search1turn38search3turn29search1turn23search0turn35view0turn35view1

At the same time, clinicians should avoid overclaiming. NICE’s 2023 stroke rehabilitation guideline did not recommend routine music therapy after stroke because the evidence was limited, heterogeneous, and lacking in cost-effectiveness data; it instead issued a specific research recommendation calling for larger pragmatic RCTs with meaningful controls. That is a good model for honest translational practice: enthusiastic about promise, disciplined about certainty. citeturn34view0turn34view1

For researchers

The field now needs fewer weakly controlled demonstrations and more mechanism-sensitive trials. Preregistration, active control conditions, equal-contact comparators, standardized biomarker timing, and assessor blinding should become routine where feasible. Receptive art-viewing research also now has emerging reporting guidance: the 2025 art-viewing review introduced the RAARR framework to improve description of settings, activities, and mechanisms. Biomarker work should use repeated sampling and multimodal outcomes because single endpoints, especially cortisol alone, are often too noisy. citeturn42search7turn21search1turn32search13turn28search0

A particularly high-value design is the hybrid lab-plus-daily-life framework: establish causal effects under controlled conditions, then test generalizability with EMA and wearables. This is also the right way to study personalization—preferred versus nonpreferred music, original versus reproduced art, solitary versus social viewing, guided versus unguided interpretation, or live versus digital performance. citeturn24search4turn32search0turn15search1turn42search8

For designers and cultural institutions

The intervention is not only the artwork; it is the encounter. The art-viewing literature suggests that setting, instructions, accessory activities, and social framing matter. Slow-looking prompts, optional reflective writing, opportunities for conversation, seating, acoustics, and sensory accessibility can change what people feel and possibly what their physiology does. Museum and gallery designers should therefore think like intervention designers, not just exhibitors. citeturn42search7turn27search2turn12search17

A trauma-informed and autonomy-supportive design stance is especially important. Some people find art calming; others find it confronting, overstimulating, or emotionally activating. Programs should therefore preserve choice, permit different engagement depths, and avoid forcing interpretation, disclosure, or social participation. This is consistent with both dementia guidance that emphasizes preference-tailoring and with the broader social prescribing principle of person-centered referral rather than one-size-fits-all programming. citeturn35view0turn33search2turn23search0

Gaps and future research agenda

The most important unresolved question is not whether art can influence mood and physiology—it can—but which ingredients drive which outcomes for which people. Current studies often package several components together: artwork content, novelty, beauty, social interaction, therapist attention, movement, expectation, and setting. Untangling these components will require dismantling trials, better controls, and formal mediation analysis. citeturn42search7turn23search0turn34view0

A second gap is the distinction between calming and engaging physiology. Art experiences can reduce cortisol and still increase electrodermal activity, heart-rate dynamics, or attentional EEG signatures. The field needs models that separate distress arousal from positive activation, awe, or “being moved.” Without that distinction, physiological findings can look contradictory when they are actually describing different layers of the same experience. citeturn42search8turn15search1turn12search17

A third gap is biomarker quality and timing. Immune and inflammatory markers remain especially underdeveloped in arts research. When they are measured, sample sizes are typically small, kinetics are not always matched to exposure windows, and confounders are substantial. More rigorous inflammatory work should use repeated measures, standardized collection windows, and realistic expectations about effect sizes. The same caution applies to cortisol, which does not track perceived stress consistently across all designs. citeturn17search2turn19search0turn21search21turn21search1

A fourth gap is ecological validity with causal rigor. Lab studies are controlled but artificial; real-world studies are realistic but noisy. The next generation of research should combine ambulatory physiology, context-aware EMA, and experimental manipulation. The fact that the EMA review already identified dozens of arts-related real-time studies means the infrastructure is now in place for this shift. citeturn32search0turn32search13turn24search4

A fifth gap is population and modality breadth. Music, visual art, and older-adult group interventions dominate the evidence base. Theater/dramatherapy, digital/interactive art, and truly multimodal combinations remain less studied, especially with biomarkers. This is not a minor omission: live-performance and synchrony work suggests that group attention and co-presence may be distinctive mechanisms, and digital/VR formats may become crucial for hospital, rural, and home-based delivery. citeturn15search20turn15search1turn12search14turn14search17

The most credible future agenda is therefore straightforward. Build adequately powered RCTs with active controls. Pair self-report with repeated autonomic, endocrine, and—where justified—immune sampling. Use portable neuroimaging and wearables when ecological realism matters. Specify activity components with reporting standards such as RAARR. Test moderators such as preference, familiarity, baseline stress, trait aesthetic responsiveness, and social context. And evaluate not only symptom change but also functioning, participation, adherence, and equity of access. If those priorities are followed, the next wave of arts-and-health research should be able to move from suggestive promise to precise, mechanism-based implementation. citeturn42search7turn21search1turn28search0turn34view0turn23search0