CoChat
Sign Up
Cited by 65Non-Invasive Vital Sign MonitoringPhysical Activity and HealthMobile Health and mHealth Applications

Response To: Investigating sources of inaccuracy in wearable optical heart rate sensors

Open access
Colvonen, Peter J.·2021·npj Digital Medicine
DOI 10.1038/s41746-021-00408-5PMC PMC7910598
Open this paper in CoChat — free accountOpen PDF

Abstract

Recently, Bent and colleagues (2020) published a timely and well-written paper examining the role of skin tone on inaccurate readings in consumer and medical grade wearables (Empatica E4 + ; Apple Watch 4; Fitbit Charge 2; Garmin Vivosmart 3; Xiaomi Miband; Biovotion Everion) 1 . They found no significant difference in accuracy across skin tones, but did find differences by devices in response to changes in activity. This finding is in contrast to previously reported studies finding wearables using green light technology had larger errors rates in tracking heart rate and energy expenditure for individuals with darker skin tones 2 , especially if exercising 3 . So while Bent and colleagues’ paper is a model in reporting in many ways, due to the incredibly important nature of this topic, it is crucial to appraise their paper to advance scientific discourse and highlight several recommendations for future researchers. Specifically, I believe their findings may be misleading due to their small sample size, which may miss important interaction effects of confounding variables and skin tone, and their use of the Fitzpatrick Skin Type Scale, which has a substantial literature of racial biases 4 , 5 , 6 , weak correlation with skin color, and large within-group variations of skin tone 4 , 7 , 8 . As such, I am concerned their findings on skin tone are not accurate and will be used to limit or misrepresent future research on inaccuracies of skin tone in wearable devices.

AI snapshot

Written from the paper’s full text.

TL;DR

This response argues that Bent et al.’s finding of no significant accuracy differences across skin tones in wearable optical heart rate sensors may be misleading because of small sample size and use of the Fitzpatrick Skin Type Scale. It recommends objective skin-tone measurement and larger samples to detect interactions with skin tone.

Key findings

  • Bent et al. reported no significant difference in accuracy across skin tones but did find differences by device in response to changes in activity; Colvonen notes this contrasts with prior studies finding green-light wearables had larger error rates in darker skin tones, especially during exercise.
  • The critique focuses on the small number of participants with the darkest skin tones: n = 9 in FST Type 6, which the author fears is too few to detect interactions between skin tone and factors such as arm hair, sweat, ambient temperature, activity, epidermal thickness, and body mass.
  • The Fitzpatrick Skin Type Scale is described as subjective, racially biased, weakly correlated with skin color, subject to large within-group variation, and inaccurate based on the administrator; the author recommends objective reflectance spectrometry/spectrocolorimetry as the new gold standard.
  • The paper recommends decreasing use of subjective skin-tone measures, increasing sample sizes to allow for interaction effects on skin tone, working with wearable companies, holding the research community accountable for bias, and including people of varying skin tones in validation and effectiveness research.

Methods

  • Design: a response/commentary arising from Bent et al. (2020), appraising their study of wearable optical heart rate sensors; no new empirical data collection or analysis is reported.
  • Evidence base: cites prior literature and studies on wearables, photoplethysmography, skin tone, pulse oximetry, and skin-tone classification; discusses Bent et al.’s use of medical-grade ECG as a reference and their reporting of confidence intervals, error rates, and missing data.
  • Main approach: argument and recommendations rather than statistical analysis.

Limitations

  • The paper does not state limitations of its own commentary; it states limitations of Bent et al., including small sample size (n = 9 in FST Type 6) and use of the Fitzpatrick Skin Type Scale.
  • Evident from the text: as a commentary, it presents no new data or formal analysis, and some claims are framed as hypotheses (“I hypothesize,” “I fear”) and rely on cited literature.
  • Evident from the text: it does not provide an alternative empirical test of skin-tone accuracy in wearables.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1038/s41746-021-00408-5

Cites

19 of 19 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 36 cited sources checked · 0 supported · 0 flagged

  • Premisecitation
    § Response To: Investigating sources of inaccuracy in wearable optical heart rate sensors

    Previously reported studies found that wearables using green light technology had larger error rates in tracking heart rate and energy expenditure for individuals with darker skin tones, especially if exercising.

    “This finding is in contrast to previously reported studies finding wearables using green light technology had larger errors rates in tracking heart rate and energy expenditure for individuals with darker skin tones2, especially if exercising3.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [2] Shcherbina, A. et al. Accuracy in wrist-worn, sensor-based measurements of heart rate and energy expenditure in a diverse cohort. J. Personalized Med. 7, 3 (2017). source

      Not checked in the free run.

    • cites [3] Fallow, B. A., Tarumi, T. & Tanaka, H. Influence of skin type and wavelength on light wave reflectance. J. Clin. Monit. Comput. 27, 313–317 (2013). source

      Not checked in the free run.

  • Premisecitation
    § Response To: Investigating sources of inaccuracy in wearable optical heart rate sensors

    Only a small handful of studies examine skin tone and accuracy rates directly.

    “there are only a small handful of studies that examine skin tone and accuracy rates directly1–3,11”
    Quote in the paperSubstantial claim0 of 4 sources checked
    • cites [1] Bent, B., Goldstein, B. A., Kibbe, W. A. & Dunn, J. P. Investigating sources of inaccuracy in wearable optical heart rate sensors. NPJ Digit. Med. 3, 1–9 (2020). source

      Not checked in the free run.

    • cites [2] Shcherbina, A. et al. Accuracy in wrist-worn, sensor-based measurements of heart rate and energy expenditure in a diverse cohort. J. Personalized Med. 7, 3 (2017). source

      Not checked in the free run.

    • cites [3] Fallow, B. A., Tarumi, T. & Tanaka, H. Influence of skin type and wavelength on light wave reflectance. J. Clin. Monit. Comput. 27, 313–317 (2013). source

      Not checked in the free run.

    • cites [11] Feiner, J. R., Severinghaus, J. W. & Bickler, P. E. Dark skin decreases the accuracy of pulse oximeters at low oxygen saturation: the effects of oximeter probe type and gender. Anesth. Analg. 105, S18–S23 (2007). source

      Not checked in the free run.

  • Premisecitation
    § Response To: Investigating sources of inaccuracy in wearable optical heart rate sensors

    Objective reflectance spectrometry accurately identifies skin color/tone using multiple color wavelengths for classification.

    “I recommend replacing it with objective reflectance spectrometry which accurately identifies skin color/tone using multiple color wavelengths for classification17”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [17] Pershing, L. K. et al. Reflectance spectrophotometer: the dermatologists’ sphygmomanometer for skin phototyping? J. Invest. Dermatol. 128, 1633–1640 (2008). source

      Not checked in the free run.

  • Premisereasoning
    § Response To: Investigating sources of inaccuracy in wearable optical heart rate sensors

    The Fitzpatrick Skin Type Scale is a subjective scale.

    “the FST is a subjective scale that classifies six skin type categories according to the amount of skin pigmentation and skin’s reaction to sun exposure.”
    • Also stated in: § Response To: Investigating sources of inaccuracy in wearable optical heart rate sensors · hedged
      “the use of the subjective FST may not accurately classify skin tone based on the administrator.”
    Quote in the paperSubstantial claim
  • Interpretationreasoning
    § Response To: Investigating sources of inaccuracy in wearable optical heart rate sensors

    Bent et al.'s conclusions may be misleading because too few people with the darkest skin tones were included.

    “I fear that their conclusions may be misleading as too few people with the darkest skin tones were included (n = 9 in FST Type 6)”
    Quote in the paperSubstantial claim
  • May combine two claims. 5
  • Classified as probably not a substantial claim. 4

  • May combine two claims.

    Premisecitation
    § Response To: Investigating sources of inaccuracy in wearable optical heart rate sensors

    Bent et al. found no significant difference in accuracy across skin tones, but did find differences by devices in response to changes in activity.

    “They found no significant difference in accuracy across skin tones, but did find differences by devices in response to changes in activity.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [1] Bent, B., Goldstein, B. A., Kibbe, W. A. & Dunn, J. P. Investigating sources of inaccuracy in wearable optical heart rate sensors. NPJ Digit. Med. 3, 1–9 (2020). source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation
    § Response To: Investigating sources of inaccuracy in wearable optical heart rate sensors

    Spectrocolorimetry generally uses multiple variables for categorizing skin tone, the most common being lightness/brightness, red/green value, and a blue/yellow value that more accurately represents empirical values of color tones.

    “Spectrocolorimetry generally uses multiple variables for categorizing skin tone. The most common variables are lightness/brightness (a gray scale from pure white to pure black), red/green value, and a blue/yellow value that more accurately represents empirical values of color tones18.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [18] Del Bino, S. & Bernerd, F. Variations in skin colour and the biological consequences of ultraviolet radiation exposure. Br. J. Dermatol. 169, 33–40 (2013). source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning
    § Response To: Investigating sources of inaccuracy in wearable optical heart rate sensors

    A lack of accurate information about error in diverse skin tones may cause unintended consequences by limiting access to accurate health information based on skin tone and reinforcing existing healthcare disparities.

    “A lack of accurate information about error in diverse skin tones may cause unintended consequences by limiting access to accurate health information based on skin tone and reinforcing existing healthcare disparities.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation
    § Response To: Investigating sources of inaccuracy in wearable optical heart rate sensors

    Several factors influence Photoplethysmography (PPG) accuracy and may cause an interaction with skin tone, including arm hair, sweat, ambient temperature, level of activity, thickness of skin epidermis, and body mass.

    “There are several factors that influence PPG accuracy that may cause an interaction with skin tone, including the presence of arm hair, sweat, ambient temperature, level of activity, thickness of skin epidermis20, and body mass21.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [20] Moço, A. V., Stuijk, S. & de Haan, G. Skin inhomogeneity as a source of error in remote PPG-imaging. Biomed. Opt. Express 7, 4718–4733 (2016). source

      Not checked in the free run.

    • cites [21] Castaneda, D., Esparza, A., Ghamari, M., Soltanpur, C. & Nazeran, H. A review on wearable photoplethysmography sensors and their potential future applications in health care. Int. J. Biosens. Bioelectron. 4, 195 (2018).

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation
    § Response To: Investigating sources of inaccuracy in wearable optical heart rate sensors

    The Fitzpatrick Skin Type Scale was originally used to assess the propensity for skin to burn, and only later became a means of describing skin tone.

    “Ware and colleagues (2020) point out that the FST was originally used to assess the propensity for skin to burn, and only later became a means of describing skin tone.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [4] Ware, O. R., Dawson, J. E., Shinohara, M. M. & Taylor, S. C. Racial limitations of fitzpatrick skin type. Cutis 105, 77–80 (2020).

      Not checked in the free run.

  • May combine two claims.

    Premisecitation
    § Response To: Investigating sources of inaccuracy in wearable optical heart rate sensors

    The Fitzpatrick Skin Type Scale was developed in 1975 by individuals with white skin for individuals with white skin.

    “Developed in 1975 by individuals with white skin for individuals with white skin6”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [6] Pichon, L. C. et al. Measuring skin cancer risk in African Americans: is the Fitzpatrick Skin Type Classification Scale culturally sensitive. Ethn. Dis. 20, 174–179 (2010).

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation
    § Response To: Investigating sources of inaccuracy in wearable optical heart rate sensors

    Bent and colleagues ran a power analysis to address sample size.

    “While Bent and colleagues ran a power analysis to address sample size”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [1] Bent, B., Goldstein, B. A., Kibbe, W. A. & Dunn, J. P. Investigating sources of inaccuracy in wearable optical heart rate sensors. NPJ Digit. Med. 3, 1–9 (2020). source

      Not checked in the free run.

7 of 7

  • The paper relies on this without support.

    Wearables are transitioning from consumer goods into health-related research and their internal algorithms are becoming FDA approved.

    “Further, these devices are now transitioning from consumer goods into health-related research and their internal algorithms are becoming FDA approved.”
1 of 1

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • The skin color of the subject has been a major point of discussion concerning the quality of PPG recordings.

    “Prominently, the skin color of the subject has been a major point of discussion concerning the quality of PPG recordings14,47,48.”

    From Data quality evaluation in wearable monitoring (2022) · cites this paper as [47]

  • These pulse oximeters have probably been calibrated on a population sample with light-skin pigmentation, which miss relevant interaction effects of confounding variables and skin pigmentation.

    “Besides, these pulse oximeters have probably been calibrated on a population sample with light-skin pigmentation, which miss relevant interaction effects of confounding variables and skin pigmentation [23,33,109].”

    From Skin pigmentation influence on pulse oximetry accuracy: a systematic review and bibliometric analysis (2022) · cites this paper as [33]

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
Colvonen, P. J.
Affiliations
University of California San Diego (US)
Corresponding author
Colvonen, Peter J.
ORCID
Colvonen, Peter J.
Venue
npj Digital Medicine
Publisher
Springer Science and Business Media LLC
Published
2021-02-26
Volume
4
Issue
1
Pages
38-38
ISSN
2398-6352 (ISSN-L)
DOI
10.1038/s41746-021-00408-5
PMC
PMC7910598
OpenAlex
W3130956469
Cited by
65
References
19
Citations by year
2026: 6 · 2025: 8 · 2024: 14 · 2023: 18 · 2022: 14 · 2021: 5
FWCI
3.87
Type
journal article
Language
en
Open access
Licence
Locations
  • npj Digital Medicine (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Topics
  • Physical Sciences › Engineering › Biomedical Engineering › Non-Invasive Vital Sign Monitoring
  • Health Sciences › Medicine › Physiology › Physical Activity and Health
  • Health Sciences › Health Professions › General Health Professions › Mobile Health and mHealth Applications
Keywords
skin tone, racial bias, measurement error, confounding variables

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-09.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-09.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review