Original article CED

Clinical and Experimental Dermatology

Skin photorejuvenation effects of light-emitting diodes (LEDs): a

comparative study of yellow and red LEDs in vitro and in vivo
S. K. Kim,1 H. R. You,2 S. H. Kim,2 S. J. Yun,2 S. C. Lee2 and J. B. Lee2
1
Namak Oracle Dermatology Clinic, Gwangju, Jeollanam-do, Korea; and 2Department of Dermatology, Chonnam National University Medical School,
Gwangju, Korea

doi:10.1111/ced.12902

Summary Background. Red-coloured light-emitting diodes (LEDs) can improve skin photo-
rejuvenation and regeneration by increasing cellular metabolic activity.
Aim. To evaluate the effectiveness of visible LEDs with specific wavelengths for skin
photorejuvenation in vitro and in vivo.
Methods. Normal human dermal fibroblasts (HDFs) from neonatal foreskin were
cultured and irradiated in vitro by LEDs at different wavelengths (410–850 nm) and
doses (0–10 J/cm2). In vivo experiments were performed on the skin of hairless mice.
Expression of collagen (COL) and matrix metalloproteinases (MMPs) was evaluated
by semi-quantitative reverse transcription PCR (semi-qRT-PCR), western blotting and
a procollagen type I C-peptide enzyme immunoassay (EIA). Haematoxylin and eosin
and Masson trichrome stains were performed to evaluate histological changes.
Results. In HDFs, COL I was upregulated and MMP-1 was downregulated in
response to LED irradiation at 595
2 and 630
8 nm. In the EIA, a peak result
was achieved at a dose of 5 J/cm2 with LED at 595
2 nm. In vivo, COL I synthesis
was upregulated in a dose-dependent manner to both 595 and 630 nm LED irradia-
tion, and this effect was prolonged to 21 days after a single irradiation with a dose
of 100 J/cm2. These histological changes were consistent with the results of semi-
qRT-PCR and western blots.
Conclusion. Specific LED treatment with 595
2 and 630
8 nm irradiation
was able to modulate COL and MMPs in skin, with the effects persisting for at least
21 days after irradiation. These findings suggest that yellow and red LEDs might be
useful tools for skin photorejuvenation.

Introduction treatment and prevention of photoageing.3 Both ablative

and nonablative therapies have been described for skin
Photoageing of skin results in reduced collagen synthe-
rejuvenation. However, ablative therapies damage the
sis and increased expression of matrix metallo-
epidermis and, thus, require a prolonged recovery time.
proteinases (MMPs).1 These alterations in collagen
Nonablative therapies have increased in popularity and
organization produce skin laxity and wrinkling.2 There-
are designed to confine selective thermal injuries to the
fore, increasing collagen production while reducing
reticular dermis, leading to fibroblast activation and syn-
MMP expression represents a potential strategy for
thesis of new collagen and extracellular matrix material
Correspondence: Dr Jee-Bum Lee, Department of Dermatology, Chonnam without any epidermal damage. Multiple devices incor-
National University Medical School, 42 Jebongro, Donggu, Gwangju, porating intense pulsed light, pulsed dye lasers, 532 nm
501-757, Korea potassium–titanyl–phosphate (KTP) lasers, and lasers at
E-mail: [email protected] various infrared wavelengths have been used.4
Conflict of interest: the authors declare that they have no conflicts of Light-emitting diode (LED) photomodulation has
interest. been described as a process for improving aged skin,
Accepted for publication 23 December 2015 which modifies cell activity using light sources,

798 Clinical and Experimental Dermatology (2016) 41, pp798–805 ª 2016 British Association of Dermatologists
 Comparative study of skin photorejuvenation effects of yellow and red LEDs  S. K. Kim et al.

without thermal effects.5 In vitro studies revealed that Semi-quantitative reverse transcription polymerase
wavelengths at near-infrared regions of the visible chain reaction
spectrum induce collagen synthesis with concurrent
Total mRNA was isolated (RNeasy Mini Kit; Qiagen,
reduction in expression of MMPs.3 Red light, in partic-
Valencia, CA, USA). Subsequently, cDNA was reverse-
ular, stimulates fibroblast proliferation.6,7
transcribed (Omniscript RT Kit; Qiagen) from 500 ng
In this study, we investigated the effectiveness and
of total RNA and subjected to semi-quantitative
longevity of LED therapy for photorejuvenation by
reverse transcription PCR (semi-qRT-PCR) (HiPi PCR
evaluating the expression levels of procollagens and
PreMixl ELPIS, Daejeon, Korea). Table 1 shows the
MMPs in human dermal fibroblasts (HDFs) in vitro and
primer sequences and thermal cycling conditions.
in murine skin in vivo after irradiation with various
Band intensities were quantified by densitometry and
LEDs at different wavelengths and doses.
normalized to b-actin or GAPDH values.

Materials and methods

Western blotting

Light-emitting diode light source Western blotting was performed as described previ-
ously.8 Protein bands were probed with rabbit antibod-
LEDs at wavelengths of 410
10, 480
7, 525
4,
ies against procollagen type-I (Y-18) MMP-1 (H-300)
580
4, 595
2, 630
8 and 850
3 nm were
(both Santa Cruz Biotechnology, Santa Cruz, CA,
used. Each LED irradiation dose was measured using a
quantum photoradiometer (DO 9721) connected to an
ultraviolet (UV)A probe (Sonda LP 9021) (both Delta
OHM, Padua, Italy). The distance from the LED module
Table 1 Primer sense and anti-sense sequences.
to the cells and mice was 100 mm.
PCR product
Gene Direction Primer sequence (50 ?30 ) size, bp
Animals
Human
Female SKH1 hairless mice 6–8 weeks old (Orient Bio COL I Forward atgatgagaaatcaaccgga 487
Co. Ltd., Seongnam, Korea) were housed in an air- Reverse ccagtagcaccatcatttcc
conditioned room maintained at 22
1 °C and COL III Forward cctccaactgctcctactcg 536
Reverse tcgaagcctctgtgtccttt
50
5% humidity, with a 12-h light/dark cycle. After
MMP-1 Forward agatgtggagtgcctgatgt 378
a 1-week stabilization period, the mice were randomly Reverse tgcaacacgatgtaagttgt
divided into three groups of animals. The study was MMP-2 Forward gctggagaaccaaagtctga 459
approved by the ethics committee of Chonnam Reverse cagtgccctcttgagacagt
National University Medical School, and performed in TGF-b1 Forward gggactatccacctgcaaga 124
Reverse cggagctctgatgtgttgaa-
accordance with the usual standards of animal care.
TIMP-1 Forward acccccgccatggagagtgt 319
Reverse gaggcaggcaggcaaggtga
GAPDH Forward gtcttcaccaccatggagaaggc 400
Human dermal fibroblast culture
Reverse cggaaggccatgccagtgagctt
Primary, normal HDF cultures were established from Mouse
Col I Forward ctttgcttcccagatgtcctat 327
explanted neonatal foreskin and used for up to 10 pas-
Reverse cggtgtcccttcattccag
sages. Cell cultures were maintained in Dulbecco mod- Col III Forward ccaccccgaactcaagagtgg 443
ified Eagle medium (DMEM; BioWhittaker Cambrex Bio Reverse ccatcctctagaactgtgtaagtg
Sciences, Walkersville, MD, USA), supplemented with Mmp-1 Forward gacttctacccatttgatgg 307
10% fetal bovine serum and 2 mmol/L glutamine. Reverse ttagggttggggtcttcatc
Mmp-2 Forward agaaaagattgacgctgtgt 396
Reverse cttcacgctcttgagacttt
Cell viability Tgf-b1 Forward tgacgtcactggagttgtacgg 169
Reverse ggttcatgtcatggatggtgc
Cell viability was assayed using a colorimetric b-actin Forward tcatgaagtgtgacgttgacatccgt 284
3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl-tetrazolium Reverse cctagaagcacttgcggtgcacgatg
bromide (MTT) kit (Chemicon International Inc., Bil- COL, collagen; GADPH, glyceraldehyde 3-phosphate dehydrogen-
lerica, MA, USA), according to the manufacturer’s ase; MMP, matrix metalloproteinase; TIMP, tissue inhibitor of
instructions. metalloproteinase; TGF, transforming growth factor.

ª 2016 British Association of Dermatologists Clinical and Experimental Dermatology (2016) 41, pp798–805 799
Comparative study of skin photorejuvenation effects of yellow and red LEDs  S. K. Kim et al.

USA), MMP-2 and anti b-actin (ab-6276, 1 : 5000; Upregulation of collagens and downregulation of
Abcam Inc., Cambridgeshire, UK) overnight at 4 °C. matrix metalloproteinases in human dermal fibroblasts

Significant increases in the mRNA levels of COL-I (type

Procollagen type-I C-peptide enzyme immunoassay I collagen a1) and COL-III and decreases in the mRNA
levels of MMP-1 and MMP-2 were observed at wave-
Procollagen I protein levels were determined using a
lengths of 595
2 and 630
8 nm administered at
commercially available, procollagen type I C-peptide
a dose of 2.5 J/cm2 (Fig. 2a).
enzyme immunoassay (EIA) kit (TaKaRa, Shiga,
On western blots, type I procollagen protein expres-
Japan). Absorbance at 450 nm was determined using
sion was significantly higher relative to normal con-
a microplate reader (Molecular Devices, Sunnyvale,
trols (NCs) (Fig. 2b), whereas MMP-1 expression was
CA, USA).
significantly decreased at wavelengths of 580
4,
595
2 and 630
8 nm. However, no statistically
Effects of light-emitting diode irradiation on murine significant changes in MMP-2 expression were
skin according to dose and elapsed time observed (Fig. 2b).

A phototest was conducted in hairless mice to investi-

gate whether LED irradiation would induce any Quantitative determination of procollagen type I
adverse reactions. LEDs at different wavelengths were C-peptide in human dermal fibroblasts
used to irradiate the dorsal murine skin at various Type I procollagen synthesis was increased in HDFs at
doses (25, 50, 100, 150 and 200 J/cm2), and the wavelengths of 580
4, 595
2 and 630
8 nm.
effects were observed over time. Mice were monitored The most prominent response was observed at a dose
for adverse responses for up to 28 days after LED of 5 J/cm2 and wavelength of 595
2 nm, with
irradiation. levels of type I procollagen synthesis reaching levels
> 100 ng/mL higher than those observed in control
Histological examination cells (Fig. 2c).

Tissue samples were collected from dorsal mouse skin

at various time points after irradiation. Samples were Light-emitting diodes modulate collagen and matrix
embedded in paraffin wax, cut into slices 4 lm thick, metalloproteinase protein mRNA and protein levels in
and stained with haematoxylin and eosin (H&E) or murine skin
Masson trichrome (MT) for histological analysis. Murine skin was treated with wavelengths of
580
4, 595
2 and 630
8 nm, which were
Statistical analysis
selected based on the in vitro experiments. No adverse
reactions were observed in murine skin following LED
Student t-test was used for the statistical analyses, irradiation (525
4, 580
4, 595
2 and
which were performed with Sigma Plot software (Sys- 630
8) at doses of 50–200 J/cm2.
tat Software, Inc., San Jose, CA, USA). Values are Expression levels of COL-I and COL-III mRNA
shown as mean
SD. P < 0.05 was considered statis- increased in a dose-dependent manner relative to the
tically significant. control at a wavelength of 595
2 nm, peaking at
100 J/cm2, and reached a plateau (Fig. S1a). The pro-
tein levels of procollagen I also increased and MMP-1
Results
decreased in a LED dose-dependent manner up to
100 J/cm2 (Fig. S1b). Following once-, twice- and
Effect of light-emitting diodes on cell survival
thrice-weekly irradiation at a dose of 100 J/cm2, the
Cell viability was evaluated after LED irradiation at vari- levels of COL-I and COL-III mRNA were significantly
ous doses (1, 2.5, 5, 10, 20 and 40 J/cm2) and wave- upregulated, whereas the levels of MMP-1 mRNA were
lengths (410
10, 480
7, 525
4, 580
4, significantly downregulated compared with control
595
2, 630
8 and 850
3 nm). Almost no levels (Fig. 3a). Although the levels of COL-I and COL-
effects were observed; however, there was reduced cell III mRNA also increased according to the number of
viability at wavelengths of 410
10 and 480
7 nm irradiation sessions, these differences were not statisti-
at a dose of 20 J/cm2 (Fig. 1). cally significant.

800 Clinical and Experimental Dermatology (2016) 41, pp798–805 ª 2016 British Association of Dermatologists
 Comparative study of skin photorejuvenation effects of yellow and red LEDs  S. K. Kim et al.

Figure 1 Cells were irradiated with various light-emitting diodes at visible to near-infrared wavelengths of 410
10, 480
7,
525
4, 580
4, 595
2, 630
8 and 850
3 nm. Cell viability was assessed via 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl-
tetrazolium bromide assay. Reduced cell viability (red rectangles) at wavelengths of 410
10 and 480
7 nm at a dose of 20 J/cm2
was seen.

In addition, western blotting revealed increased pro- over time. Marked histological changes were observed
tein levels of procollagen 1 in response to wavelengths in skin stained with H&E at days 16 and 21 after LED
of 595
2 and 630
8 nm. Production of COL-I irradiation with 595
2 and 630
8 nm. MT stain
remained consistent, regardless of the number of appli- showed more densely deposited collagen in the dermis
cations. These results (Fig. 3b) were consistent with of LED-irradiated skin compared with control skin
those of mRNA expression (Fig. 3a). After LED irradia- (Fig. 4c).
tion at 595
2 and 630
8 nm (dose 100 J/cm2),
upregulated levels of COL-I, COL-III and transforming
Discussion
growth factor (TGF)-b mRNA and downregulated
levels of MMP-1 mRNA were observed in a time- Skin ageing is caused by both chronological ageing
dependent manner from day 1 to day 21. In particu- and photoageing, and these independent, clinically
lar, COL-I was strongly expressed at later time points and biologically distinct processes simultaneously
after LED treatment (from day 16; P < 0.05) (Fig. 4a). affect the skin. However, decreased collagen synthesis
The effects of LED treatment persisted up to day 21. In is observed in both intrinsic and extrinsic ageing pro-
western blots, COL-I expression increased and MMP-I cesses.9 Thus, various rejuvenation techniques, partic-
expression decreased over time. Levels of COL-I protein ularly LED devices, have been used in an attempt to
increased consistently along with the mRNA expres- promote collagen biosynthesis with a concurrent
sion levels (Fig. 4b). reduction in expression of MMPs.3,10–12 Lam et al.
demonstrated that irradiation of fibroblasts with
633 nm wavelength light increased procollagen pro-
Light-emitting diodes induce collagen deposits in the
duction by an average of four-fold, without affecting
murine dermis
the activities of collagenase and gelatinase, which con-
Tissue samples from dorsal mouse skin were stained trol collagen degradation.7 Numerous clinical studies
with H&E and MT to evaluate histological changes have shown the efficacy of irradiation at 633 and

ª 2016 British Association of Dermatologists Clinical and Experimental Dermatology (2016) 41, pp798–805 801
Comparative study of skin photorejuvenation effects of yellow and red LEDs  S. K. Kim et al.

Figure 2 Human dermal fibroblasts after treatment with light-emitting diodes (LEDs), showing (a) expression of collagen and matrix
metalloproteinases (MMPs) in human dermal fibroblasts; (b) mRNA levels of collagen (COL)-1 and COL-III, MMP-1 and MMP-2, trans-
forming growth factor-b (TGF-b), and tissue inhibitors of metalloproteinases-1 (TIMP-1), assessed via semi-quantitative reverse tran-
scription PCR, and protein levels of COL-I, MMP-1 and -2 assessed via western blotting; and (c) procollagen I protein levels determined
by enzyme immunoassay (EIA). Bars indicate mean
SEM. *P < 0.05, **P < 0.01 vs. normal control (NC).

Figure 3 (a) mRNA and (b) protein levels

of collagen (COL) and matrix metallopro-
teinase (MMP) were determined by semi-
quantitative reverse transcription PCR
(semi-qRT-PCR) and western blotting,
respectively, which in murine skin were
found to be independent of the number
of applications of light-emitting diode
irradiation at 580
4, 595
2 and
630
8 nm.

660 nm for skin rejuvenation.13–15 Russel et al. periorbital wrinkles in 81% of patients at the 12-week
reported that the combined use of 633 and 830 nm follow up.14 Lee et al. reported that photodynamic
wavelengths led to significant improvements in therapy, using wavelengths of 613–645 nm,

802 Clinical and Experimental Dermatology (2016) 41, pp798–805 ª 2016 British Association of Dermatologists
 Comparative study of skin photorejuvenation effects of yellow and red LEDs  S. K. Kim et al.

Figure 4 (a,b) Murine dermal cells irradiated with various light emitting diodes (LEDs) had changes in (a) mRNA expression of collagen
(COL)-I and COL-III, matrix metalloproteinase (MMP)-1, MMP-2 and transforming growth factor (TGF)-b, as determined by semi-qRT-
PCR, and (b) COL-I and MMP-1 protein levels over time, as determined by western blotting. Bars indicate mean
SEM. *P < 0.05,
**P < 0.01 vs. normal control. (c) Histological studies showed that collagens were more densely deposited in the dermis of LED-irra-
diated skin than in the dermis of controls. (Upper panels) Haematoxylin and eosin, original magnification 9200; (lower panels) Masson
trichrome, original magnification 9200.

ª 2016 British Association of Dermatologists Clinical and Experimental Dermatology (2016) 41, pp798–805 803
Comparative study of skin photorejuvenation effects of yellow and red LEDs  S. K. Kim et al.

significantly induced the expression of the pro- adenosine triphosphate production occurs rapidly after
inflammatory cytokines, IL-1b, TNF-a and TGF-b.16 LED photomodulation, and subsequently triggers meta-
Furthermore, in a recent study, expression of Fos, the bolic activity in fibroblasts.10,20
key molecular component contributing to DNA dam- In the present study, the effects of 595 nm LED irra-
age after exposure to UV light, decreased after treat- diation might have been mediated through an increase
ment with red light (670 nm).17 in cytochromes, a primary target in fibroblast mito-
As several reports have shown that red LEDs are chondria, leading to subsequent collagen remodelling,
effective for rejuvenation both in vivo and in vitro, we as has been shown with yellow light at ~590 nm.
designed a series of experiments to investigate the
effectiveness and longevity of LED irradiation at speci-
Conclusion
fic wavelengths on the expression of collagen and
MMP. After exposing HDF to various LED wave- We have demonstrated that both yellow and red LEDs
lengths, collagen and MMP expression levels were are useful and effective for promoting collagen synthe-
assessed by semi-qRT-PCR and western blot analyses. sis in vitro and in vivo. Therefore, LEDs with specific
Our findings revealed that light emitted at wave- wavelengths can be an effective, safe and promising
lengths of 595 and 630 nm concurrently upregulated tool for skin rejuvenation.
expression of COL-I and COL-III and downregulated
expression of MMP-1 and MMP-2 in vitro (Fig. 2a). In
addition, procollagen I protein levels were increased by Acknowledgements
the same wavelengths of light (Fig. 2b), peaking at This research was financially supported by the Min-
595 nm at a dose of 5 J/cm2 (Fig. 2c). istry of Knowledge Economy (MKE), Korea Institute for
To determine the extent and duration of skin photo- Advancement of Technology (KIAT) and Honam Insti-
rejuvenation, LEDs at 580, 595 and 630 nm were tute for Regional Program Evaluation through the
used to irradiate mice at different frequencies and Leading Industry Development for Economic Region.
doses. COL-I mRNA and protein levels increased in a
dose-dependent manner (Fig. S1a,b), but were not
dependent on the number of applications (Fig. 3a,b).
Regarding the longevity of the LED effects, both 595 What’s already known about this topic?
and 630 nm LEDs were used to administer a single
● Photomodulation improves ageing skin,
100 J/cm2 dose to murine skin. The effects on COL-I
improving skin quality by stimulating fibroblast
mRNA and protein levels persisted for 21 days after
proliferation.
LED irradiation (Fig. 4a,b). This study demonstrated
that following a single exposure to LED, the positive
effect on collagen synthesis was maintained for at least
3 weeks. Significant COL-I upregulation was observed
in the 595 nm yellow LED-irradiated group relative to What does this study add?
that of the 630 nm red LED-irradiated group (Fig. 4b). ● We evaluated the effects of LED photomodula-
Yellow LED light was also reportedly effective in tion on expression of procollagens and MMPs in
other studies. Weiss et al. demonstrated the efficacy of human dermal fibroblasts and murine skin.
a yellow LED at a wavelength of approximately ● We showed that yellow and red LEDs promote
590 nm,10,18 and reported that a regimen of eight collagen synthesis in vitro and in vivo for up to
590 nm LED treatments over 4 weeks led to an aver- 21 days after irradiation, suggesting that these
age 28% increase in density of collagen, with 4% LEDs are effective and safe tools for skin rejuve-
reduction in MMP-1 levels.10 In addition, McDaniel nation.
et al. reported that levels of markers of extracellular
signal-regulated kinase and cytokine pathways, such
as c-Jun and c-Fos, were reduced by treatment with a
590 nm LED array.19
Regarding the mechanism of LED photomodulation,
References
direct absorption of light molecules by fibroblast mito- 1 Fisher GJ, Kang S, Varani J et al. Mechanisms of
chondria has been proposed; this would lead to photoaging and chronological skin aging. Arch Dermatol
increased cell activity and collagen production. New 2002; 138: 1462–70.

804 Clinical and Experimental Dermatology (2016) 41, pp798–805 ª 2016 British Association of Dermatologists
 Comparative study of skin photorejuvenation effects of yellow and red LEDs  S. K. Kim et al.

2 Fisher GJ, Wang ZQ, Datta SC et al. Pathophysiology of Omnilux Revive facial treatment in skin rejuvenation.
premature skin aging induced by ultraviolet light. N Engl Lasers Med Sci 2005; 20: 6–10.
J Med 1997; 337: 1419–28. 14 Russell BA, Kellett N, Reilly LR. A study to determine the
3 Barolet D. Light-emitting diodes (LEDs) in dermatology. efficacy of combination LED light therapy (633 nm and
Semin Cutan Med Surg 2008; 27: 227–38. 830 nm) in facial skin rejuvenation. J Cosmet Laser Ther
4 Nelson JS, Majaron B, Kelly KM. What is nonablative 2005; 7: 196–200.
photorejuvenation of human skin? Semin Cutan Med Surg 15 Barolet D, Roberge CJ, Auger FA et al. Regulation of skin
2002; 21: 238–50. collagen metabolism in vitro using a pulsed 660 nm LED
5 Weiss RA, McDaniel DH, Geronemus RG. Review of light source: clinical correlation with a single-blinded
nonablative photorejuvenation: reversal of the aging study. J Invest Dermatol 2009; 129: 2751–9.
effects of the sun and environmental damage using laser 16 Choi JY, Park GT, Na EY et al. Molecular changes
and light sources. Semin Cutan Med Surg 2003; 22: following topical photodynamic therapy using methyl
93–106. aminolaevulinate in mouse skin. J Dermatol Sci 2010;
6 Young S, Bolton P, Dyson M et al. Macrophage 59: 198–203.
responsiveness to light therapy. Lasers Surg Med 1989; 9: 17 Myakishev-Rempel M, Stadler I, Polesskaya O et al. Red
497–505. light modulates ultraviolet-induced gene expression in
7 Ts L, Rp A, Ca M et al. Laser stimulation of collagen the epidermis of hairless mice. Photomed Laser Surg 2015;
synthesis in human skin fibroblast cultures. Lasers Life Sci 33: 498–503.
1986; 1: 61–77. 18 Weiss RA, McDaniel DH, Geronemus RG et al. Clinical
8 Choi JY, Piao MS, Lee JB et al. Propionibacterium acnes experience with light-emitting diode (LED)
stimulates pro-matrix metalloproteinase-2 expression photomodulation. Dermatol Surg 2005; 31: 1199–205.
through tumor necrosis factor-a in human dermal 19 McDaniel DH, Newman J, Geronemus R et al. Non-
fibroblasts. J Invest Dermatol 2008; 128: 846–54. ablative non-thermal LED photomodulation – a
9 Varani J, Warner RL, Gharaee-Kermani M et al. Vitamin multicenter clinical photoaging trial. Lasers Surg Med
A antagonizes decreased cell growth and elevated 2003; 15: 22.
collagen-degrading matrix metalloproteinases and 20 McDaniel DH, Weiss RA, Geronemus R et al. Light-tissue
stimulates collagen accumulation in naturally aged interactions I: photothermolysis vs photomodulation
human skin. J Invest Dermatol 2000; 114: 480–6. laboratory findings. Lasers Surg Med 2002; 14: 25.
10 Weiss RA, Mcdaniel DH, Geronemus RG et al. Clinical
trial of a novel non-thermal LED array for reversal of
photoaging: clinical, histologic, and surface profilometric Supporting Information
results. Lasers Surg Med 2005; 36: 85–91. Additional Supporting Information may be found in
11 Lee SY, Park KH, Choi JW et al. A prospective,
the online version of this article:
randomized, placebo-controlled, double-blinded, and split-
Figure S1. Light-emitting diodes modulate collagen
face clinical study on LED phototherapy for skin
rejuvenation: clinical, profilometric, histologic,
and matrix metalloproteinase protein mRNA and pro-
ultrastructural, and biochemical evaluations and tein levels in murine skin. (a) Expression levels of col-
comparison of three different treatment settings. lagen (COL)-I and COL-III mRNA increased and those
J Photochem Photobiol, B 2007; 88: 51–67. of matrix metalloproteinase (MMP)-1 decreased in a
12 Baroulet D, Boucher A, Bjerrring P. In vivo human dose-dependent manner relative to the normal controls
dermal collagen production following LED-based therapy: at a wavelength of 595
2 nm. (b) Protein levels of
the importance of treatment parameters. Lasers Surg Med COL I, MMP-1 shown by western blotting. Bars indi-
2005; 17(Suppl): 76. cate mean
SEM. *P < 0.05, **P < 0.01 vs. normal
13 Bhat J, Birch J, Whitehurst C et al. A single-blinded control (NC).
randomised controlled study to determine the efficacy of

ª 2016 British Association of Dermatologists Clinical and Experimental Dermatology (2016) 41, pp798–805 805