At first glance, the rise of wearable health technology seems like a democratizing force in healthcare. These devices promise to empower individuals to monitor their health metrics without frequent clinical visits. Common devices include fitness trackers like Fitbit and Apple Watch, continuous glucose monitors, and smart rings that track sleep patterns. However, beneath this promise lies a troubling reality. Many of these devices fail to serve marginalized communities effectively. This is particularly true for individuals with darker skin tones. This failure creates what I term a “digital health divide.”

The Science Behind the Disparity

The core issue stems from the basic technology these devices use. Most wearable devices rely on photoplethysmography (PPG) (Kim & Baek, 2023). PPG is a simple optical technique. It uses light sensors to measure various health metrics by detecting blood flow patterns beneath the skin. The device shines green or red LED light into the skin. The light that bounces back helps measure heart rate and blood oxygen levels.

However, melanin creates a significant problem for this technology. Melanin is the pigment that gives skin its color (Morison, 1985). It absorbs light at many of the same wavelengths used by PPG sensors. This interference leads to less accurate readings. Research by Shcherbina et al. studied this issue in detail, focusing on seven commercially available fitness trackers, including the Apple Watch, Basis Peak, Fitbit Surge, Microsoft Band, Mio Alpha 2, PulseOn, and Samsung Gear S2 (Shcherbina et al., 2017). They found that darker skin tones, which contain higher levels of melanin, were associated with greater errors in heart rate measurements across devices. This issue arises because melanin absorbs light at the wavelengths used by photoplethysmography (PPG) sensors, reducing the accuracy of signal detection. Among the tested devices, the Apple Watch demonstrated the lowest overall error for heart rate measurements, while the Samsung Gear S2 had the highest error, highlighting differences in device design and sensor sensitivity.

The accuracy problems don’t stop at heart rate monitoring. Blood oxygen sensors, which became essential during the COVID-19 pandemic for detecting hypoxia, also exhibit similar biases. A study by the US Food and Drug Administration (FDA) and other researchers found that pulse oximeters tend to overestimate oxygen saturation levels (SpO2) in individuals with darker skin tones (Feiner et al., 2007). This inaccuracy becomes more pronounced as actual oxygen saturation (SaO2) decreases, particularly when SpO2 drops below 80%. These findings raise critical concerns about the reliability of pulse oximetry in diverse populations, as overestimating oxygen levels in patients with pigmented skin could delay necessary medical interventions, potentially endangering their health. It could mean the difference between identifying and missing a serious medical condition.

Socioeconomic Barriers to Access

Technical limitations are just one part of the issue. Socioeconomic barriers further widen the gap. While wearables are becoming more common and useful in clinical settings, lower-income households often lack access to these devices and the necessary digital infrastructure (Holko et al., 2022). Without addressing this disparity, digital health tools risk becoming yet another factor contributing to unequal health outcomes.

The cost barrier extends beyond the initial purchase. Many devices require:

• Smartphone compatibility for data tracking
• Stable internet connection for data sync
• Regular charging capabilities
• Replacement bands or accessories
• Subscription fees for advanced features

These ongoing costs create additional barriers for economically disadvantaged communities.

Algorithmic Bias and Data Interpretation

The challenges with wearable health devices extend beyond just hardware limitations to include the software that processes and interprets the data. The algorithms responsible for analyzing data from these devices often carry significant biases. These biases typically arise because the datasets used to train these algorithms are often limited or unrepresentative, particularly lacking diversity in race, ethnicity, and skin tones (Merid & Volpe, 2023). As a result, devices may misinterpret or inaccurately analyze health data for individuals from marginalized racial and ethnic groups. 

This issue is particularly concerning in areas like heart rate and blood oxygen detection, where algorithms may perform poorly for individuals with darker skin tones. The consequences of these biases are not just technical errors; they can lead to missed diagnoses, delayed treatments, and worsened health outcomes for underrepresented groups. Instead of helping to reduce health disparities, biased algorithms can inadvertently reinforce them, making it crucial for future innovations to address these issues with greater diversity and inclusivity in their design and testing.

Real-World Health Implications

The inaccuracies in wearable health technology have serious consequences. For instance, unreliable heart rate readings may cause a Black athlete to receive incorrect training recommendations. In another case, a dark-skinned patient’s continuous glucose monitor may consistently underreport blood sugar levels. These errors led to inappropriate medication dosing.

The situation becomes even more problematic when considering conditions that disproportionately affect marginalized communities. Consider these statistics:

• Diabetes affects Black Americans at nearly twice the rate of white Americans (Factors contributing to higher incidence of diabetes for black, 2018)
• Heart disease rates are significantly higher in minority communities (Heart disease and African Americans, n.d.)
• Sleep disorders are often underdiagnosed in communities of color (Billings et al., 2021)

These are precisely the health conditions that wearable technology claims to help monitor. When these devices fail to work accurately for the communities most at risk, they effectively widen existing health disparities.

Moving Toward Inclusive Innovation

Addressing these disparities requires a comprehensive approach. Here are several key steps to consider:

1. Diversity in Development Teams
   1. Include researchers and engineers from various racial and ethnic backgrounds
   2. Incorporate feedback from diverse user groups during early development stages
   3. Establish partnerships with community health organizations
2. Improved Testing Standards
   1. Require testing across all skin tones before device approval
   2. Implement standardized accuracy metrics for different user populations
   3. Conduct real-world testing in diverse communities
3. Technical Solutions
   1. Develop new sensor technologies less affected by melanin
   2. Create algorithms trained on diverse datasets
   3. Design devices with adjustable settings for different skin tones

Policy and Regulatory Changes

Current regulatory frameworks need updating. The FDA’s approval process for wearable health devices lacks specific requirements for diverse testing. Here are several proposed policy changes:

• Mandatory diversity requirements in clinical trials
• Transparency in reporting accuracy rates across different populations
• Regular post-market surveillance of device performance in diverse communities
• Financial incentives for developing more inclusive technologies

Future Directions and Hope

Despite current challenges, recent innovations show promise. New technologies are emerging that may help bridge the digital health divide. For example:

• Radio frequency sensors that don’t rely on light penetration (Rajiv, 2023)
• Community-based testing and development programs

Conclusion

The promise of wearable health technology can only be realized through inclusive design and implementation. These devices must serve all communities effectively. As they become increasingly integrated into healthcare delivery, addressing their limitations is crucial. This is not just a technical challenge but an ethical imperative. The future of digital health must prioritize equity and accessibility. Only then can technological advances help narrow, rather than widen, existing health disparities.

References

Billings, M. E., Cohen, R. T., Baldwin, C. M., Johnson, D. A., Palen, B. N., Parthasarathy, S., Patel, S. R., Russell, M., Tapia, I. E., Williamson, A. A., & Sharma, S. (2021). Disparities in Sleep Health and Potential Intervention Models: A Focused Review. Chest, 159(3), 1232–1240. https://doi.org/10.1016/j.chest.2020.09.249.

Factors contributing to higher incidence of diabetes for black. (2018b, January 23). National Institutes of Health (NIH). https://www.nih.gov/news-events/nih-research-matters/factors-contributing-higher-incidence-diabetes-black-americans#:~:text=In%20the%20U.S.%2C%20black%20adults,over%20the%20last%2030%20years.

Feiner, J. R., Severinghaus, J. W., & Bickler, P. E. (2007). Dark skin decreases the accuracy of pulse oximeters at low oxygen saturation: the effects of oximeter probe type and gender. Anesthesia & Analgesia, 105(6), S18-S23. https://journals.lww.com/anesthesia-analgesia/fulltext/2007/12001/dark_skin_decreases_the_accuracy_of_pulse.4.aspx. 

Heart disease and African Americans. (n.d.). Office of Minority Health. https://minorityhealth.hhs.gov/heart-disease-and-african-americans#:~:text=In%202019%2C%20African%20Americans%20were,to%20non%2DHispanic%20white%20women.

Holko, M., Litwin, T. R., Munoz, F., Theisz, K. I., Salgin, L., Jenks, N. P., Holmes, B. W., Watson-McGee, P., Winford, E., & Sharma, Y. (2022). Wearable fitness tracker use in federally qualified health center patients: strategies to improve the health of all of us using digital health devices. NPJ digital medicine, 5(1), 53. https://doi.org/10.1038/s41746-022-00593-x.

Kim, K. B., & Baek, H. J. (2023). Photoplethysmography in wearable devices: a comprehensive review of technological advances, current challenges, and future directions. Electronics, 12(13), 2923. https://doi.org/10.3390/electronics12132923. 

Merid, B., & Volpe, V. (2023). Race Correction and Algorithmic Bias in Atrial Fibrillation Wearable Technologies. Health equity, 7(1), 817–824. https://doi.org/10.1089/heq.2023.0034. 

Morison, W. L. (1985). What is the function of melanin?. Archives of dermatology, 121(9), 1160-1163. 

Rajiv. (2023, September 22). Radio Frequency Sensors: Exploring types and applications. RF Page. https://www.rfpage.com/radio-frequency-sensors-types-applications/. 

Shcherbina, A., Mattsson, C. M., Waggott, D., Salisbury, H., Christle, J. W., Hastie, T., … & Ashley, E. A. (2017). Accuracy in wrist-worn, sensor-based measurements of heart rate and energy expenditure in a diverse cohort. Journal of personalized medicine, 7(2), 3. https://doi.org/10.3390/jpm7020003.