White light designed around lens yellowing eased color discrimination for volunteers over 70
Fudan University and OPPLE Lighting report 26 of 34 volunteers aged 70 and up discriminated colors better under P1 white light than simulated daylight. The model screened 11,541 candidate spectra that met standard white-light and color-rendering requirements near 5000 K before P1 and a second design were chosen for testing.
The Scientist · Science desk
Two part-of-whole bars. Under the P1 spectrum, 26 of 34 participants (76.5%) improved color discrimination. Under the P2 spectrum, 22 of 34 participants (64.7%) improved.
- Improved under P1, claim 11 26 of 34 participants 76.5%
Improved under P1 The rest of the 34
- Improved under P2, claim 11 22 of 34 participants 64.7%
Improved under P2 The rest of the 34
| Share | Part | Whole | Percent | Claim |
|---|---|---|---|---|
| Improved under P1 | 26 participants | 34 participants | 76.5% | 11 |
| Improved under P2 | 22 participants | 34 participants | 64.7% | 11 |
What happened
- The model estimates how easily someone can tell similar colors apart from the lighting spectrum, the wavelengths objects reflect, and age-adjusted sensitivities of the eye's three cone types.
- Of the two spectra chosen, P1 was tuned for discrimination and P2 balanced it against natural color; for a modeled 85-year-old they were predicted to recover about 49% and 36% of the age-related loss.
- The 34 volunteers, all 70 or older, did a color-sorting test under P1, P2 and the reference light, with illuminance and color temperature nearly identical across the three.
- Both optimized spectra significantly improved sorting performance over the simulated daylight reference.
Why it matters
- capability A designer working inside standard white-light limits can rank candidate spectra by predicted benefit to older eyes, because spectra that looked similar differed substantially in that prediction.
- decision Anyone specifying lamps for a shared room has to pick a point between P1 and P2. P2 was designed to favor natural-looking color and showed less benefit than P1 in both the model and the volunteers.
- constraint A fixed spectrum helps most older users but not all of them: 8 of 34 did not improve under P1 and 12 of 34 did not improve under P2.
- precedent Dai said the same approach could be extended to how displays appear to older viewers, so the model's logic may be tested outside room lighting.
The problem is optical. Ordinary color-quality scoring misses it. "As people age, the lens inside the eye gradually yellows and absorbs more short-wavelength light, which can make similar colors harder to discriminate," said research team leader Qi Dai of Fudan University [1]. Conventional ways to measure color quality generally use a mathematical representation of human color vision that does not account for how sensitivity to different wavelengths changes with age [2].
Two sets of numbers in this study answer different questions. The first set is the model's prediction for a hypothetical 85-year-old. By our calculation the predicted recoveries leave roughly 51% of the modeled loss in place under P1 and 64% under P2 [12][13]. The second set is what people did. The volunteers were all 70 or older, and the summary does not give their age spread [9].
The two sets agree on order. P1 helped 76.5% of participants and P2 64.7%, or 22 of 34 [11]. Agreement in rank supports the model's ranking of the two designs. It does not validate the absolute percentages, because the volunteers sorted colors and the model predicts recovery for a single modeled age.
The control is a reasonable one for the question asked. Holding illuminance and color temperature nearly identical across the three conditions removes brightness and overall tint as easy explanations for a sorting difference, which leaves the shape of the spectrum [9].
On the other half of the design goal, how the light looks to everyone else, the evidence is compliance. The spectra met conventional criteria for white-light chromaticity and color rendering [16], and the researchers report that familiar colors still looked natural [18]. "Since the spectra also met conventional criteria for white-light chromaticity and color rendering, they could be used for lighting spaces shared by people of different ages," Dai said [17].
The summary does not report the size of the sorting gain, or whether viewers with typical color vision judged these lights. "Significantly improved" is the only effect description given [10].
What to watch
- A follow-up that reports the size of the sorting gain and includes viewers with typical color vision under the same lights.
- A comparison of P1 against lamps already installed in homes or care facilities, rather than against simulated D50 daylight.
- Any display work that applies the model to how screens look to older viewers, as Dai described.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence58
- Adoption
- Insufficient
- Hype gap+15
- Incentives40
- Confidence55
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
"As people age, the lens inside the eye gradually yellows and absorbs more short-wavelength light, which can make similar colors harder to discriminate," said research team leader Qi Dai from Fudan University in China.
- [2]
Conventional ways to measure color quality generally use a mathematical representation of human color vision that does not account for how the eye's sensitivity to different wavelengths changes with age due to lens yellowing.
- [3]
Researchers from Fudan University and OPPLE Lighting report in Optics Express that, in adult volunteers aged 70 and older, lighting spectra designed with the new model improved color discrimination compared with an LED-simulated D50 daylight reference at nearly the same color temperature.
- [4]
The model estimates how easily a person can distinguish similar colors based on the illumination light's spectrum, the wavelengths reflected by colored objects, and age-adjusted sensitivities of the eye's three cone types.
- [5]
The researchers evaluated 11,541 possible light spectra that met standard requirements for white light and color rendering near 5000 K, a neutral, daylight-like white.
- [6]
Spectra that otherwise appeared similar could differ substantially in how they were predicted to affect older adults' ability to distinguish between similar colors.
- [7]
The researchers selected two spectra: P1, optimized to improve color discrimination, and P2, designed to balance color discrimination with natural-looking colors.
- [8]
For a modeled 85-year-old observer, P1 was predicted to recover about 49% of the age-related loss in color discrimination while P2 recovered about 36%.
- [9]
34 adult volunteers aged 70 and above performed a color-sorting test under lighting with the P1 and P2 spectra and an LED-simulated daylight reference; illuminance and color temperature were nearly identical in all three conditions.
- [10]
Both optimized spectra significantly improved performance on the color-sorting test compared with the simulated daylight reference.
- [11]
At the individual level, 26 of 34 participants (76.5%) improved color discrimination with the P1 spectrum, and 22 of 34 (64.7%) improved with P2.
- [12]
If P1 is predicted to recover about 49% of the modeled loss, roughly 51% remains.
- [13]
If P2 is predicted to recover about 36% of the modeled loss, roughly 64% remains.
- [14]
8 of 34 participants did not improve color discrimination under P1.
- [15]
12 of 34 participants did not improve color discrimination under P2.
- [16]
The spectra also met conventional criteria for white-light chromaticity and color rendering.
- [17]
"Since the spectra also met conventional criteria for white-light chromaticity and color rendering, they could be used for lighting spaces shared by people of different ages," said Dai.
- [18]
In the volunteers aged 70 and older, familiar colors still looked natural under the optimized lighting.
- [19]
Dai said the approach could be expanded to capture how displays appear to older viewers.
Sources
1 independent publisher whose own reporting we read for this story.
- phys.orgComputational model guides lighting design to improve color discrimination in aging eyes
1 article · October 13, 2026
Topics and entities
Follow any of these and your For You feed starts watching them — no settings page required.
Entities
- Optics ExpressFollow
- Fudan UniversityFollow
- Qi DaiFollow
- OPPLE LightingFollow