Molecules and Photoproducts Generated in Human Skin by the Four Major Solar Wavebands

Claude AI Deep Research June 2026

TL;DR

• Each solar waveband generates a distinct chemistry in skin: UVB drives a now-greatly-expanded family of 7-dehydrocholesterol (7-DHC) photoproducts (previtamin D3/vitamin D3, plus lumisterol3 and tachysterol3 and their CYP11A1/CYP27A1 hydroxy-derivatives), DNA photoproducts (CPDs, 6-4PPs), cis-urocanic acid, and POMC-derived peptides (β-endorphin, α-MSH); UVA mobilizes nitric oxide from cutaneous nitrite/nitrosothiol stores and generates ROS/HO-1; visible/blue light acts through OPN3 to drive melanogenesis and (therapeutically) isomerizes bilirubin; infrared-A acts on mitochondrial cytochrome c oxidase (photobiomodulation) but also drives MMP-1/photoaging.
• The strongest proven benefits are narrow: vitamin D3 synthesis (mechanism definitive; supplement-outcome RCTs largely null except cancer-mortality signals), neonatal bilirubin phototherapy (definitive RCT-grade), and red/NIR photobiomodulation for wrinkles/wound healing (multiple RCTs). The Slominski lumisterol/tachysterol photoprotection work and the Weller/Feelisch UVA-NO blood-pressure work are mechanistically strong but clinically still preliminary (small/short human studies, mostly in vitro/ex vivo or single-exposure).
• Every beneficial pathway carries a tradeoff: UVB CPDs are the dominant carcinogenic lesion; UVA causes "dark CPDs," photoaging and is only a weak BP-lowering tool in the one (underpowered) RCT; blue light causes durable hyperpigmentation in skin of color; IR-A degrades dermal collagen. Net public-health framing should weigh non-burning, moderate exposure rather than either avoidance or excess.

Key Findings

1. UVB is the most chemically productive band and the dominant carcinogenic one

Its signature beneficial reaction — photoconversion of 7-DHC to previtamin D3 — peaks at ~295–300 nm (action-spectrum maximum ~297–298 nm; CIE 174:2006), essentially nil above 315 nm. The same band produces the carcinogenic CPDs and 6-4 photoproducts, cis-urocanic acid (immunosuppression), and POMC peptides.

2. The "expanded photoproducts of 7-DHC" (Slominski et al.) are real, endogenous, and biologically active but non-calcemic

UVB over-irradiation converts previtamin D3 to lumisterol3 (L3) and tachysterol3 (T3); CYP11A1 and CYP27A1 then hydroxylate these (and 7-DHC and D3 directly) into a large family of hydroxy-metabolites detected in human serum and epidermis. These act largely outside the classical genomic VDR — on RORα/γ (as inverse agonists), AhR, LXRα/β, and PPARγ — and show photoprotective, anti-inflammatory, anti-proliferative and pro-differentiation activity. Evidence tier: strong biochemistry + in vitro/ex vivo human skin; in vivo clinical efficacy not yet established.

3. UVA's headline benefit is nitric-oxide mobilization lowering blood pressure (Weller/Feelisch)

UVA photo-releases NO from cutaneous nitrite and S-nitrosothiol stores independent of NO synthase, causing vasodilation and a transient BP fall. This is mechanistically well-supported and reproduced, but the only randomized hypertension trial (2-week daily low-dose UVA) showed only a transient post-irradiation clinic-BP drop and no effect on 24-hour ambulatory BP — and was severely underpowered (n=13 of a planned 80).

4. Visible/blue light works through skin opsins, and bilirubin phototherapy is the one unambiguous visible-light clinical win

OPN3 in melanocytes senses blue light (~415–450 nm) and drives Ca²⁺/CaMKII→MITF melanogenesis, explaining durable hyperpigmentation in Fitzpatrick III–VI. Blue light (460–490 nm; optimum ~478 nm) isomerizes unconjugated bilirubin to excretable photoisomers (lumirubin), the basis of neonatal jaundice phototherapy.

5. Infrared-A is double-edged: photobiomodulation benefit vs. photoaging harm via the same mitochondrial target

Red/NIR (~630–850 nm) light absorbed by cytochrome c oxidase boosts ATP, modulates ROS/NO and supports RCT-backed skin rejuvenation and wound healing; yet IR-A also triggers retrograde mitochondrial ROS signaling that upregulates MMP-1 (collagen breakdown) in human skin in vivo.

Details

I. UVB (~280–315 nm)

A. The vitamin D photosynthetic cascade (proven mechanism)

The B-ring of 7-DHC absorbs UVB, breaking the C9–C10 bond to form previtamin D3, which thermally isomerizes over hours to vitamin D3 (cholecalciferol). The action spectrum (MacLaughlin & Holick, Science 1982; codified as CIE 174:2006) maxima at ~297 nm with negligible production >315 nm; Holick's group (Science 1980) established the photosynthesis in human skin. A 2021 PNAS in-vivo re-test (Young, Harrison et al., PNAS 2021, using serum 25(OH)D3 endpoints) proposed a systematic correction, finding the original action spectrum's longer wavelengths somewhat over-weighted. Vitamin D3 → 25(OH)D3 (liver) → 1,25(OH)2D3 (kidney and skin). - Proven benefits: prevention/treatment of rickets/osteomalacia and support of calcium/bone homeostasis (definitive). - Contested/possible benefits: The VITAL RCT (Manson et al., NEJM 2019; 25,871 participants, 2000 IU/d D3) found no reduction in primary endpoints of invasive cancer (HR 0.96, 95% CI 0.88–1.06) or major CVD (HR 0.97, 0.85–1.12), nor all-cause mortality (HR 0.99, 0.87–1.12). A secondary signal suggested reduced cancer mortality (HR 0.83, 0.67–1.02) in analyses excluding early follow-up, and reduced advanced/metastatic cancer especially at normal BMI. Autoimmune-disease reduction was a later VITAL secondary finding. The bulk of extra-skeletal supplement RCT evidence is null. - Harms/tradeoffs: none from cutaneous synthesis itself (it is self-limiting — excess previtamin D3 and D3 photodegrade); the UVB needed carries CPD/cancer risk.

B. Lumisterol3 (L3), tachysterol3 (T3) and their CYP-derived hydroxy-metabolites (Slominski program)

Continued UVB on previtamin D3 yields L3 and T3. Once considered inert, both are now substrates for steroidogenic CYPs: - CYP11A1 converts L3 → 20(OH)L3, 22(OH)L3, 24(OH)L3, 20,22(OH)2L3 (Tuckey et al., Int J Biochem Cell Biol 2014); converts T3 → 20S(OH)T3 (Slominski et al., FASEB J 2022); and acts on 7-DHC → 22(OH)7DHC, 20,22(OH)2-7DHC, 7-dehydropregnenolone. The side chain of L3/7-DHC can also be cleaved by CYP11A1 to 7-dehydropregnenolone (7DHP) or pregna-lumisterol (pL). - CYP27A1 converts L3 → 25(OH)L3, 27(OH)L3 (Tuckey et al., 2018) and T3 → 25(OH)T3 (Slominski et al., FASEB J 2022). - Endogenous concentrations (human): lumisterol3 serum ~19.2 ± 2.7 ng/mL and epidermis 2.5 ± 1.1 ng/mg protein; tachysterol3 serum 7.3 ± 2.5 ng/mL (n=4) and epidermis 25.1 ± 5.2 ng/mg protein; for comparison epidermal 7-DHC 227.3 ± 59.4 and vitamin D3 0.2 ± 0.05 ng/mg protein (Slominski et al., FASEB J 2022; J Invest Dermatol 2024 review). Hydroxylumisterols [20(OH)L3, 22(OH)L3, 24(OH)L3] were detected in serum and epidermis (Slominski et al., Sci Rep 2017;7:11434), with epidermal mono-hydroxylumisterol levels significantly exceeding parental L3 (p<0.01) and serum 20(OH)L3 ~9× higher than serum 20(OH)D3; 20S(OH)T3 and 25(OH)T3 are detected endogenously but could not be reliably quantified. - Receptor targets: hydroxylumisterols do not bind the genomic VDR pocket; they act as inverse agonists of RORα/γ (20(OH)L3 being the most potent RORγ inhibitor), and on the non-genomic VDR site, plus AhR, LXRα/β and PPARγ. Tachysterol-hydroxyderivatives act on VDR (genomic, stimulating CYP24A1 ~10× weaker than 1,25(OH)2D3), AhR (strong agonism, 20S(OH)T3 > 25(OH)T3), LXRα/β (coactivator-recruitment EC50 ~10⁻⁸–10⁻⁷ M) and PPARγ (IC50 ~10⁻⁷ M for 20S(OH)T3, ~10⁻⁶ M for 25(OH)T3) (Slominski et al., FASEB J 2022;36(8):e22451). - Photoprotective activity (in vitro/ex vivo): 24-hydroxylumisterol3 reduced UV-induced CPDs and oxidative (8-OHdG) DNA damage in human keratinocytes and human skin explants (p<0.001 in explants), comparable to 1,25(OH)2D3, via a glycolysis-dependent enhancement of nucleotide-excision repair (XPC/XPA upregulation; effect abolished by 2-deoxyglucose and by XPC/XPA knockdown) (De Silva et al., Metabolites 2023;13:775). Hydroxylumisterols suppressed UVB-induced NF-κB activation and inflammatory cytokines (IL-17, IFN-γ, TNF-α) and promoted keratinocyte differentiation (Chaiprasongsuk et al., IJMS 2020; Nrf2/p53-mediated antioxidant defense in Redox Biol 2019). Anti-fibrogenic effects are RORγ-dependent (Janjetovic et al., Endocrinology 2021). - Evidence tier: strong biochemistry and cell/explant pharmacology; no human clinical efficacy trials yet; in-vivo photoprotection demonstrated for related compounds in Skh:hr1 mice but 24(OH)L3 itself tested only in vitro/ex vivo.

C. DNA photoproducts: CPDs and 6-4 photoproducts (proven harm)

Direct absorption of UVB by adjacent pyrimidines forms cyclobutane pyrimidine dimers (CPDs) and pyrimidine(6-4)pyrimidone photoproducts (6-4PPs), repaired by nucleotide excision repair. CPDs account for ≥80% of UVB-induced mutations in mammalian cells (You et al., JBC 2001) and drive the UV-signature C→T transitions in skin cancers. There is no proven physiological benefit; CPDs are the principal initiating lesion of keratinocyte cancers and melanoma. (Notably, CPDs are also formed by UVA — see below.)

D. Urocanic acid (trans→cis isomerization)

trans-Urocanic acid in the stratum corneum isomerizes to cis-UCA on UVB; the action spectrum for cis-UCA production peaks at 280–310 nm, red-shifted vs. the erythema spectrum. cis-UCA mediates UV-induced immunosuppression (proposed to act via the 5-HT2A receptor; Walterscheid et al., PNAS 2006). - Possible benefit: trans-UCA may provide weak endogenous UV screening and stratum-corneum pH/barrier homeostasis; immunosuppression could theoretically dampen some autoimmune/photodermatosis activity. - Harm: immunosuppression contributing to photocarcinogenesis.

E. POMC-derived peptides: β-endorphin, α-MSH, ACTH (proven generation, mixed significance)

UVB-induced DNA damage activates keratinocyte p53 → transcription of proopiomelanocortin (POMC), cleaved to α-MSH (drives MC1R→MITF→eumelanin tanning), ACTH, and β-endorphin (Cui et al., Cell 2007). - β-endorphin: skin generation confirmed in human keratinocytes in vivo after narrowband UVB (Exp Dermatol 2016, Kemeny group); proposed basis of "UV addiction" (Fell et al., Cell 2014, mouse). β-endorphin also suppresses UVB-induced epidermal barrier damage via NF-κB/mTORC1 (Kim et al., Sci Rep 2023). - α-MSH/eumelanin tanning is genuinely photoprotective (facultative pigmentation), though delayed and modest (the SPF equivalent of a tan is low). - Evidence tier: generation proven; "addiction"/mood benefits speculative in humans.

F. UVB ROS, and the vitamin D photoprotection loop

UVB generates reactive oxygen species and oxidative DNA damage. Notably, locally produced 1,25(OH)2D3 reduces UV-induced CPDs, increases p53, lowers nitric oxide products, and reduces photoimmunosuppression in keratinocytes, mice and (topically) human skin via a non-genomic VDR/ERp57 pathway (Mason group: Gupta et al. 2007; Sequeira et al. 2012; Damian et al. 2010 topical calcitriol human study). Tradeoff: topical calcitriol reduced UV genetic damage but suppressed cutaneous immunity in humans (Damian et al., Exp Dermatol 2010).

II. UVA (~315–400 nm)

A. Nitric oxide mobilization (Weller/Feelisch) — strong mechanism, weak clinical effect so far

Human skin holds large enzyme-independent NO stores (nitrite, S-nitrosothiols/nitroso-proteins), greatly exceeding circulating NO. UVA photodecomposes these to bioactive NO (Mowbray et al.; Oplander et al., Circ Res 2009 — whole-body UVA lowered BP via NO release). Liu, Feelisch, Weller et al. (J Invest Dermatol 2014;134:1839) showed two SED of UVA to 24 volunteers lowered BP with decreased circulating nitrate and increased nitrite, independent of NOS and independent of vitamin D; dietary nitrate manipulation didn't alter the effect. - Proven (mechanistic) effect: acute UVA → cutaneous NO release → arterial vasodilation → transient BP fall. - Clinical trial: the only randomized hypertension trial (Weller et al., J Hum Hypertens 2023, PMC10328825) was a sham-controlled crossover in n=13 with high-normal/stage-1 hypertension (130–159/85–99 mmHg), 5 J/cm² full-body UVA (320–410 nm) daily × 14 days. It was underpowered — the investigators "aimed to recruit 80 such patients to give us adequate power to detect a 2 mmHg fall in ABP" but enrolled only 13. Result: a transient clinic-BP fall (−8.0/−3.8 mmHg vs sham, p=0.034/0.029) but no change in 24-h ambulatory BP — i.e., insufficient as a standalone antihypertensive in this small/short study. - Possible benefits: a candidate explanation for latitude/season BP and CVD-mortality gradients; the "sunlight benefits beyond vitamin D" hypothesis. Epidemiology (Lindqvist MISS cohort) is supportive but observational and confounded. - Harms: UVA is the chief driver of photoaging and contributes to skin cancer (see B).

B. ROS / singlet oxygen and "dark CPDs" (harm)

UVA is poorly absorbed by DNA directly; it generates ROS and singlet oxygen via endogenous photosensitizers, producing 8-oxo-guanine and, predominantly at TT sites, CPDs (Mouret et al., PNAS 2006). Premi et al. (Science 2015;347(6224):842–7) showed UVA-activated NADPH oxidase and NO synthase generate peroxynitrite that chemiexcites melanin to a triplet state: "CPDs are generated for >3 hours after exposure to UVA…These 'dark CPDs' constitute the majority of CPDs and include the cytosine-containing CPDs that initiate UV-signature C→T mutations," most prominent with pheomelanin. This reframes melanin as partly photosensitizing and links UVA to melanoma mutagenesis.

C. Melanin photo-oxidation / immediate pigment darkening (IPD)

UVA oxidizes preexisting melanin → immediate (transient) and persistent pigment darkening; photoprotective value is minimal.

D. Heme oxygenase-1 / bilirubin / ferritin (adaptive protection)

UVA-induced oxidative stress induces HO-1 in human skin fibroblasts (Tyrrell group; UVA-dependent HO-1 → ferritin increase, Vile & Tyrrell), raising ferritin and bilirubin (an endogenous antioxidant) — an adaptive cytoprotective response. HO-1 induction mediates UVA photoimmunoprotection in mice (Allanson & Reeve, J Invest Dermatol). Basis for UVA1 phototherapy of morphea/localized scleroderma (also via HO-1/MMP-1). Evidence tier: strong cell/mouse mechanism; UVA1 phototherapy is clinically used for sclerosing skin conditions.

III. Visible light (~400–700 nm)

A. Blue light (~415–450 nm), OPN3 and melanogenesis (proven in skin of color)

Regazzetti et al. (J Invest Dermatol 2018) identified OPN3 as the melanocyte blue-light sensor: 415 nm light → Ca²⁺-dependent CaMKII→CREB/ERK/p38 → MITF phosphorylation → tyrosinase/DCT, plus a tyrosinase/DCT complex causing sustained activity and long-lasting hyperpigmentation specifically in FST III–VI. Subsequent work implicates an OPN3–TRPV1 axis. - Harm: visible-light/blue-light hyperpigmentation worsens melasma and post-inflammatory hyperpigmentation in skin of color; clinically relevant because mineral SPF50+ does not block VL — iron-oxide tinted sunscreens do (multiple RCTs in melasma/skin of color). - Possible benefits: blue-light phototherapy for acne (porphyrin photoexcitation, antibacterial), and circadian/mood signaling (mostly ocular, not cutaneous). - Other opsins (OPN1-SW, OPN2/OPN4/OPN5) are reported in skin at mRNA level but their cutaneous function is less established; flavin-mediated ROS contribute to blue-light damage.

B. Red light (~630–700 nm) — cytochrome c oxidase / photobiomodulation

Overlaps with IR-A mechanism (Section IV); red light photons are absorbed by cytochrome c oxidase (complex IV) and water/ion channels. Clinical: red/NIR LED RCTs show wrinkle reduction and increased collagen density — Wunsch & Matuschka (Photomedicine and Laser Surgery 2014;32(2):93–100; PMID 24286286), a controlled trial of 136 participants concluding that "both novel light sources…have demonstrated efficacy and safety for skin rejuvenation and intradermal collagen increase when compared with controls." Note: blue light is excluded from the PBM definition by some authors due to genotoxic/non-photobiomodulatory effects.

C. Bilirubin photoisomerization — neonatal jaundice phototherapy (definitive clinical use)

Unconjugated 4Z,15Z-bilirubin in skin absorbs blue light maximally ~460 nm; therapeutic window 460–490 nm (optimum ~478 nm), irradiance ≥30 µW/cm²/nm. Reactions: rapid configurational isomerization (→ 4Z,15E, the major immediate, reversible product), slower structural isomerization to lumirubin (Z-lumirubin, ~5%, the main excreted species), and photo-oxidation. Products are water-soluble and excreted without hepatic conjugation. This is the one visible-light therapy with unequivocal RCT-grade clinical benefit (prevents kernicterus; AAP 2024 technical report).

D. Visible-light melanogenesis as photoadaptation

Visible light induces melanogenesis through a photoadaptive response; single exposures give little, repeated exposures durable pigmentation (Mahmoud et al., J Invest Dermatol 2010).

IV. Infrared (700 nm–1 mm), especially IR-A (~760–1400 nm)

A. Cytochrome c oxidase, ATP, photobiomodulation (benefit) — contested primary chromophore

Red-to-NIR photons (action peaks ~660–680 and ~810–830 nm matching CCO absorption) are absorbed by cytochrome c oxidase; the proposed mechanism is photodissociation of inhibitory NO from CCO, restoring electron transport, increasing mitochondrial membrane potential and ATP, with modulated ROS and intracellular Ca²⁺ activating redox-sensitive transcription (Hamblin, Photochem Photobiol 2018; Wong-Riley et al.; Karu). - Caveat/contest: some work shows PBM proliferation effects persist in cells lacking CCO (Lima et al., J Photochem Photobiol 2019), and Hamblin & Liebert (2022) describe mechanisms "beyond CCO" (light/heat-gated ion channels, interfacial water). - Proven benefits (RCTs): skin rejuvenation/wrinkle reduction and collagen increase (Wunsch & Matuschka 2014); wound healing; oral mucositis (well-supported). - Possible benefits: broader PBM claims (neuro, musculoskeletal) — variable evidence.

B. NO release from CCO

As above — IR/red light displaces NO from cytochrome c oxidase; this NO is both a signaling molecule and locally vasoactive.

C. Heat shock proteins

IR/heat induces HSPs (e.g., HSP70/HSP27); proposed cytoprotection/preconditioning (preclinical).

D. Mitochondrial ROS and retrograde signaling → MMP-1 (harm)

IR-A penetrates deeply to the dermis; physiological doses generate mitochondrial ROS triggering a retrograde (mitochondria→nucleus) signaling response that upregulates MMP-1 (collagen-degrading) without compensatory TIMP-1, and downregulates COL1A1 — demonstrated in cultured human fibroblasts and in human skin in vivo (Schieke, Schroeder, Krutmann et al.; Schroeder et al., J Invest Dermatol 2008: 80% of subjects upregulated dermal MMP-1). This is the molecular basis for IR-A's contribution to photoaging and the rationale for antioxidant-containing "beyond-UV" photoprotection. - Tension in the field: the SAME mitochondrial target underlies both PBM benefit and IR-A photoaging harm; dose, wavelength sub-band, spectral purity (lab IR-A vs. solar IR with heat) and endpoint differ, and this dose-dependence is incompletely resolved.

Cross-cutting evidence-tier summary

• Tier 1 (definitive mechanism + clinical): vitamin D3 photosynthesis for bone health; neonatal bilirubin phototherapy; red/NIR PBM for wrinkles/wounds.
• Tier 2 (strong mechanism, limited/mixed clinical): UVA-NO → BP (mechanism strong, RCT weak/underpowered); HO-1/UVA1 phototherapy for morphea; vitamin D non-genomic photoprotection (mostly mouse/ex vivo + small human topical studies); β-endorphin generation.
• Tier 3 (robust biochemistry, in vitro/ex vivo, no clinical efficacy yet): lumisterol/tachysterol hydroxy-derivative non-calcemic activities and photoprotection (Slominski program).
• Tier 4 (observational/contested): sunlight all-cause-mortality/CVD benefits beyond vitamin D (Lindqvist MISS cohort: 29,518 Swedish women, 2,545 deaths over median ~20-yr follow-up; sun avoiders had hazard ratio 2.0, 95% CI 1.6–2.5, for all-cause mortality vs the highest-exposure group, "resulting in excess mortality with a population attributable risk of 3%," J Intern Med 2014;276:77–86; dose-dependent lower CVD/non-cancer mortality, J Intern Med 2016); extra-skeletal vitamin D supplement benefits (largely null in VITAL except cancer-mortality/autoimmune secondary signals).
• Proven harms: UVB CPDs/6-4PPs (carcinogenesis), cis-UCA immunosuppression, UVA dark-CPDs/photoaging, blue-light hyperpigmentation in SOC, IR-A MMP-1 photoaging.

Recommendations

1. For the vitamin D research repository framing: Present vitamin D3 photosynthesis as mechanistically definitive but decouple it from extra-skeletal supplement efficacy claims — cite VITAL's null primary endpoints (cancer HR 0.96; CVD HR 0.97) and the cancer-mortality (HR 0.83) and autoimmune secondary signals explicitly. Threshold to revise: a positive primary-endpoint RCT or individual-participant meta-analysis (e.g., pooled D-Health/VITAL analyses).
2. Treat the Slominski lumisterol/tachysterol corpus as the most important "frontier" content but tier it clearly as in vitro/ex vivo + biochemistry. Track for (a) the first human in vivo efficacy study (photoprotection or anti-inflammatory) and (b) the absolute serum/epidermal concentrations of individual hydroxylumisterols — note that exact ng values for 20(OH)L3, 22(OH)L3, 24(OH)L3 and 20,22(OH)2L3 reside in Table 1 of Slominski et al., Sci Rep 2017;7:11434 (open access), which should be pulled directly; 20S(OH)T3 and 25(OH)T3 currently have no published quantitative concentrations.
3. For UVA-NO/cardiovascular claims: state the mechanism is real but the clinical antihypertensive effect is unproven (single small underpowered crossover RCT, transient clinic-BP only, no 24-h ABP effect). Benchmark that would change this: an adequately powered RCT (the authors' own target was ~80 patients) showing sustained 24-h ambulatory BP reduction, or hard CVD endpoints.
4. For any "sunlight benefits" public-health messaging: advocate moderate, non-burning exposure; flag that the Lindqvist sun-exposure/mortality data are observational and confounded (lifestyle, reverse causation). The dominant, proven harms (CPD carcinogenesis, photoaging) justify avoiding erythema and using broad-spectrum + iron-oxide (visible-light) protection in melasma/skin of color.
5. For photobiomodulation/red-light content: red/NIR has genuine RCT support for skin rejuvenation and wound healing; present IR-A photoaging as the dose/spectrum-dependent flip side and note the unresolved CCO-vs-beyond-CCO mechanistic debate.

Caveats

• Much of the most exciting molecular work (lumisterol/tachysterol hydroxy-derivatives, opsin signaling, dark CPDs) is in vitro, ex vivo, or in rodents; human clinical translation is largely absent.
• Action spectra are imperfect: the canonical previtamin D3 action spectrum (CIE 174:2006) has been challenged and proposed for correction (Young et al., PNAS 2021).
• The UVA-NO blood pressure literature is dominated by a small number of groups (Weller, Feelisch) with small sample sizes; one author discloses a sunscreen-company interest (Dr Weller Ltd).
• Vitamin D observational mortality associations are heavily confounded; RCT supplement data do not confirm most extra-skeletal benefits.
• IR-A findings often depend on artificial sources without the heat of solar IR; spectral sub-band and dose definitions vary across studies, limiting direct comparability, and the same mitochondrial pathway underlies both PBM benefit and photoaging harm.
• Quantitative concentration data for several T3/L3 hydroxy-metabolites are incomplete (20S(OH)T3, 25(OH)T3 detected endogenously but not reliably quantified).