A large new study tracking nearly 800 students over three months finds that consumer smartwatches—using heart rate and heart rate variability to infer “stress”—have almost no relationship with how people say they actually feel, though the devices do better at measuring sleep. The research, part of a programme aiming to build an early-warning system for depression, raises urgent questions about how Thais who use wearables should interpret stress scores, how employers and clinicians might rely on such data, and what researchers must do next to make physiological monitoring clinically useful Gizmodo The Guardian Leiden University.
Wearable makers commonly advertise stress-tracking as a core feature: Garmin, for example, says stress is “estimated by the Firstbeat Analytics engine, primarily using a combination of HR and HRV data” and presents stress as a 0–100 score Garmin. But the new peer-reviewed paper—titled “Associations between ecological momentary assessment and passive sensor data in a large student sample” and accepted for publication in the Journal of Psychopathology and Clinical Science—shows that those wrist-derived stress numbers often do not map onto real-time self-reports of emotion. The research team collected ecological momentary assessments (EMA)—short questionnaire prompts delivered to participants’ phones four times daily—while participants wore Garmin Vivosmart 4 devices continuously for up to three months, producing one of the largest datasets comparing subjective experience with consumer wearable output PubMed entry for the study.
The core finding is stark: for the majority of individuals in the sample, wearable-based physiological measures and self-reported stress showed “very weak to no associations.” One of the study’s senior authors, an associate professor in clinical psychology, told reporters the correlation between self-reported stress and the watch’s stress score was “basically zero,” and noted that heartbeat increases are not emotion-specific—heart rate rises not only with anxiety but with excitement, sexual arousal, exercise or even a surprise The Guardian Leiden University. The paper also finds that for roughly a quarter of participants, the smartwatch labeled them as stressed when they reported feeling calm, or vice versa.
Why the mismatch matters for Thai users and services is practical as well as scientific. Wearables are growing in Thailand: market reports forecast rising revenue and user penetration in the region, with fitness-tracker penetration expected to reach double digits in the near term Statista outlook for Thailand wearables. Many Thais use smartwatches for steps, sleep monitoring, and health nudges; employers and wellness programmes have also experimented with device-provided metrics. If stress scores are noisy and context-blind, people might make poor decisions—overreacting to a false “high stress” alert or ignoring psychological needs because a device signals “low stress.” That is particularly sensitive in Thailand where mental-health stigma, rising anxiety among students, and workplace pressures intersect; misinterpreted data could cause unnecessary worry or, conversely, mask people who need help.
The study sits within a growing literature that questions how much emotional information wrist-worn sensors can reliably provide. A 2023 scoping review of wearables for stress management concluded that while devices are widely studied, the effect of wearable-based approaches on reducing stress had not been definitively shown, and many studies were descriptive rather than evaluative PMC scoping review 2023. Earlier experimental work from other European groups similarly reported that consumer smartwatches often struggle to distinguish physiologically similar states such as excitement and stress Vrije Universiteit Amsterdam 2023 news release.
Methodologically, the new paper’s strength is in combining high-frequency self-report (EMA) with continuous passive sensing in a large sample (about 781 students). The wearable provided heart-rate-based metrics, a sleep duration estimate, and proprietary “body battery” or fatigue indices that manufacturers like Garmin derive from HR/HRV plus activity patterns Garmin support & blog Garmin Connect report. The researchers report a clearer relationship between the wearables and sleep duration—about two hours more recorded sleep on the device when participants later reported a better sleep night—while fatigue/body-battery scores showed a modest association. Stress, however, remained largely unexplained by the wrist sensors. The authors argue this reflects both measurement limits (optical heart-rate sensors, proprietary aggregation) and conceptual problems: stress is multifaceted, context-dependent and partly cognitive; physiological arousal alone lacks the specificity to distinguish types of arousal OSF preprint of the study PubMed.
Experts outside the study emphasise balance. A researcher from the University of Manchester noted the value of wearable data but warned against treating it as “objective truth,” urging that physiological metrics be interpreted alongside context and subjective experience The Guardian. This perspective is central to mental-health practice in Thailand, where clinicians and public-health officials increasingly combine self-report screening with behavioural data to detect depression or crisis risks. The WARN-D programme that produced the study aims to create an early-warning system for depressive episodes; the authors caution that passive data alone is unlikely to be sufficient for accurate, individualized alerts without richer contextual inputs Leiden University.
For Thai workplaces and policymakers the study raises three practical implications. First, employers should avoid using consumer stress scores as a basis for individual performance assessment, discipline, or hiring decisions: the data are noisy and susceptible to false positives and negatives, and could unfairly label employees as “stressed” when they are excited or exercising legal/privacy analysis on workplace wearables. Second, corporate wellness programmes that offer wearables should pair device data with voluntary self-reports, counselling access, and strict privacy protections; anonymised, aggregated trends may be useful at a population level, but individual-level inferences are risky. Third, national health services and NGOs should treat wearable signals as one component of screening pipelines and continue to invest in human-led outreach: Thailand already has crisis and mental-health hotlines (for example, the national Mental Health Hotline and services coordinated via the National Health Security Office) and these human-centred services remain essential for triage and care NHSO mental-health integration page WHO feature on suicide prevention in Thailand.
Culturally, Thailand offers contexts that make the wearable-stress disconnect especially visible. Public festivals, temple ceremonies, or family gatherings can produce physiological arousal that might be labelled “stress” by a watch but experienced as joy—situations where a device cannot read the social meaning of a raised heart rate. Conversely, quiet, high-stakes situations (e.g., caregiving at home, financial worry) may produce low-motion, low-heart-rate signatures that a wrist sensor underestimates, while the user experiences high psychological strain. These misalignments underline the study’s message: numbers need narratives. Thai readers should interpret device alerts with curiosity, not alarm.
What can be done to bridge the gap between sensor output and lived experience? The study and experts point to several paths: multimodal sensing that goes beyond HR/HRV (for example combining skin conductance, respiration, voice-sentiment analysis and context from phone use), better algorithms trained on labelled datasets that include behavioural context, and hybrid systems that prompt short EMAs when a device detects a spike—asking “Are you feeling stressed or excited?” in the moment. Researchers are also exploring whether population-level signals (changes in activity, sleep, social isolation) can predict depression onset even if momentary emotion labeling remains imperfect; this suggests wearables may still play a role in public mental-health surveillance and prevention if used carefully and ethically osf preprint/WARN-D project.
For Thai clinicians and researchers, the findings argue for careful validation before integrating consumer wearables into clinical workflows. University and hospital research teams should test devices in Thai populations, calibrate algorithms to local physiology and lifestyles, and combine passive sensing with culturally adapted EMA instruments. Policymakers should require transparent validation from vendors if devices are to be used in programmes tied to health advice or insurance, and should mandate clear consent, data protection and the right to opt out for employees and patients alike.
Practical recommendations for Thai smartwatch users follow directly from the evidence. First, do not let a “stress” score alone drive major decisions: use it as a prompt to reflect, not a diagnosis. Second, keep a simple diary or respond to brief EMA prompts: pairing subjective notes with device data gives a far clearer picture of patterns over time. Third, wear the device consistently (including during sleep) if you want better baseline comparisons—manufacturers themselves advise more continuous wear for improved estimates Garmin. Fourth, use the device’s sleep tracking as a more reliable signal: the new study found stronger associations between wearable-recorded sleep duration and how rested participants felt Leiden University. Fifth, if you are worried about mental health, contact trained professionals and crisis services rather than relying only on device alerts—Thailand’s network of hotlines and public mental-health services remain the appropriate entry point for care NHSO/WHO resources https://www.who.int/news-room/feature-stories/detail/suicide-prevention-in-thailand–a-whole-of-society-approach.
Looking ahead, the study is unlikely to end the wearable-stress conversation—but it should change it. Device manufacturers must be transparent about what their algorithms can and cannot do; researchers must publish open validation studies; clinicians must demand clinical-grade evidence before deploying wearables for diagnosis or treatment; and users should treat stress metrics as one instrument among many. For Thailand, where young adults are among the largest adopter groups of wearables and where mental-health services are evolving rapidly, the message is clear: wearables can enrich our view of behaviour and sleep, but they do not yet offer a reliable, context-aware readout of subjective stress. Policy, workplace practice, and personal use should reflect those limits.
The study’s authors see a constructive path forward: combine passive sensing with momentary self-report and richer contextual data to build warning systems that flag patterns (for example, sustained drops in activity and sleep) rather than single, label-based alerts. That hybrid model respects the study’s finding—physiology alone is insufficient—while harnessing wearables’ strength in continuous measurement. In the meantime, Thai readers who own smartwatches can keep using them for steps and sleep, take stress scores with a grain of salt, and prioritise direct human support when emotional or mental-health concerns arise.
Sources used in this report include the Gizmodo coverage of the new paper (Gizmodo), The Guardian’s reporting and interview with a study author (The Guardian), the Leiden University summary of the research project (Leiden University), the PubMed entry for the accepted paper (PubMed), the OSF preprint of the study (OSF preprint), Garmin’s explanation of its stress metrics (Garmin), and a 2023 scoping review on wearables for stress management (PMC review 2023). Additional context on wearables in Thailand and workplace privacy practices came from market reports and legal analyses (Statista Thailand wearables, Blank Rome on workplace wearables). For local support, readers can consult national mental-health resources and hotline services coordinated by Thai health authorities (NHSO mental-health integration, WHO on Thailand suicide prevention).