Skin color affects the effectiveness of phototherapy for neonatal jaundice

A theoretical study published in the journal Biophotonics Discovery shows that skin color and other optical properties of the skin significantly change how much therapeutic light is actually absorbed by bilirubin in the treatment of neonatal jaundice. According to the authors' calculations, as skin pigmentation increases, the proportion of light reaching the target decreases, and the optimal wavelength for phototherapy shifts - from ≈460 nm for light skin to ≈470 nm for dark skin. The conclusion is simple and inconvenient: "universal" lamps and the same irradiation modes may not work equally effectively in children of different phototypes; the spectrum and power of therapy should be adjusted to the child.

Background of the study

Neonatal jaundice is one of the most common reasons for hospitalization of newborns; the standard of treatment is phototherapy with blue/blue-green light, which converts unconjugated bilirubin into water-soluble photoisomers (including lumirubin) and thereby accelerates its elimination. Therefore, clinical guidelines emphasize a narrow effective range of wavelengths (approximately 460-490 nm) and sufficient irradiation intensity; it is in this spectral window that bilirubin absorption is maximum, and the light penetrates sufficiently deeply through the infant's tissues.

However, not all the energy emitted by the lamp reaches the “target” (bilirubin in the skin and superficial vessels): some of the light is absorbed by melanin and hemoglobin, and scattering in the multilayered skin “smears” the flow. When these optical properties change, the effective wavelength also changes: a number of studies have already hinted that blue-green light ~478-480 nm can have a stronger phototherapeutic effect than the “classic” blue peak ~460 nm, which is associated with a better balance of “bilirubin absorption ↔ penetration depth”.

A separate issue is the measurement of bilirubin by non-invasive devices (TcB): accuracy is significantly affected by skin color. In different studies, both underestimation and overestimation compared to serum bilirubin (TSB) were found in children with darker skin; recent controlled analyses and in-vitro models tend to suggest that dark skin more often leads to systematic measurement bias, and therefore high or “borderline” TcB values require confirmation by TSB.

Against this background, studies that quantitatively describe how exactly skin pigmentation and other skin properties affect the absorbed “useful” dose during phototherapy and the choice of optimal wavelength are relevant. A new study in Biophotonics Discovery solves this problem by modeling light transfer in the skin of newborns and shows that as pigmentation increases, the proportion of energy reaching bilirubin decreases, and the spectrum optimum shifts toward longer waves (from ≈460 nm to ≈470 nm). These findings fit into a broader conversation about the need to take skin color into account in optical medical technologies - from phototherapy to pulse oximetry.

How it was studied

A team from the University of Twente, Izala Hospital and UMC Groningen built computer models of how light passes through the multilayered skin of newborns and calculated how the "useful" absorbed dose of bilirubin changes under different conditions. They varied:

• Pigmentation (melanin) is the main factor that “intercepts” blue light in the epidermis;
• Hemoglobin and bilirubin content are competing absorbents that affect the depth of penetration;
• Scattering and thickness of skin layers are the parameters that determine where the light flux is "smeared".
  Modeling was performed in the entire blue range of phototherapy (about 430-500 nm), assessing at what wavelengths bilirubin absorbs maximum energy depending on the properties of the skin. The results are in close agreement with what has long been noticed in the clinic "in practice", but is rarely taken into account formally: dark skin requires a different spectral setting.

Key findings - in simple terms

The authors show three key effects: first, the darker the skin, the less "useful" light reaches bilirubin, which means that phototherapy will be slower at the same power. Second, the peak efficiency shifts: for light skin, the maximum absorbed dose of bilirubin is approximately at 460 nm, for dark skin - closer to 470 nm. Third, not only melanin "plays" the result, but also hemoglobin/bilirubin in the skin and light scattering - these are additional adjustment knobs if the device can switch the spectrum and dose. Together, this explains why the same lamps and "hourly protocols" give different rates of TcB/TSB decline in children of different phototypes.

What this changes in practice - ideas for "personalized phototherapy"

For clinics and manufacturers, the results logically lead to specific steps:

• Spectral adaptation: use sources with switchable wavelengths (e.g. combinations of blue LEDs 455-475 nm) and select the working peak taking into account the phototype.
• Dosimetry "on the skin" and not "at the lamp": focus on the absorbed dose of bilirubin, and not only on the irradiation on the mattress; ideally, use built-in sensors/models that take pigmentation into account.
• Taking into account accompanying optical factors: hemoglobin, bilirubin in the skin and scattering also change the efficiency - algorithms for adjusting the power by feedback (by the TcB/TSB dynamics) are useful.
• Correct interpretation of TcB in dark skin: devices systematically underestimate TcB in high pigmentation - it is worth confirming with serum bilirubin more often and updating calibrations.

Why This Isn't a Surprise for Biophotonics

Photonic medicine has already encountered the "skin color effect" in pulse oximetry and other optical technologies: melanin "eats" light, changing both the penetration depth and the signal-to-noise ratio. In neonatal phototherapy, this factor was long underestimated because "blue" lamps were considered universal. The new work closes the methodological gap: it qualitatively confirms the decrease in efficiency in dark skin and quantitatively shows how the optimal wavelength shifts - which provides engineering specifications for next-generation devices.

Limitations and what's next

This is a simulation, not a randomized clinical trial; numerical estimates depend on the adopted optical parameters of the skin and geometric assumptions. But the results are in good agreement with independent data: in-vitro and clinical series show underestimation of TcB and differences in response to light in children with dark skin. The next step is pilot clinical protocols with tuning LED matrices, where the spectrum/power is selected for the phototype and the rate of bilirubin reduction and the duration of hospitalization are compared.

Who is especially interested in this?

• For neonatologists and nurses - for correct interpretation of TcB and selection of intensity/duration of phototherapy in children with dark skin.
• For development engineers - for designing multispectral systems with automatic adjustment to the optical properties of the skin.
• To regulators and guideline authors - to update phototherapy standards taking into account phototype (as is already done for oximetry).

Original source: AJ Dam-Vervloet et al. Effect of skin color and other skin properties on the efficacy of phototherapy for neonatal jaundice (Biophotonics Discovery, 2025), doi: 10.1117/1.BIOS.2.3.032508.