Photodermatol Photoimmunol Photomed 1998: 14: 160–163 Copyright C Munksgaard 1998

Printed in Denmark ¡ All rights reserved

ISSN 0905-4383

Reliability of self-assessment in determining

skin phototype for Korean brown skin
Park SB, Suh DH, Youn JI. Reliability of self-assessment in determining S. B. Park, D. H. Suh, J. I. Youn
skin phototype for Korean brown skin. Photodermatol Photoimmunol Department of Dermatology, Seoul National
Photomed 1998: 14: 160–163. C Munksgaard, 1998. University College of Medicine, Seoul, Korea

The widely used skin phototype system first introduced by Fitzpatrick,

which is based on the history of an individual tendency to burn and
tan, was studied in 733 Korean men by means of a self-assessment ques-
tionnaire. Combining two factors, individual burning and tanning ten-
dency, less than half (43.9%) could be classified according to the skin
phototype system. Furthermore, the concordant ratio between subjective
skin phototypes assessed by the subjects themselves and classification by
combining self-reported burning and tanning tendencies was much lower
(34.4%). In a subgroup of 156 subjects, the minimal erythema dose Key words: burning and tanning tendency; skin
(MED) was determined and compared with the reported tendencies to burn phototype; brown skin
and tan. The results showed that subjects with higher MEDs tended to Jai-Il Youn, Department of Dermatology,
report more tanning. We concluded that in Korean brown skin, there is Seoul National University College of Medicine,
a disconcordant relationship between reported burning and tanning ten- 28 Yongon-dong, Chongno-gu, Seoul 110–744,
dencies and, though reliable for white skin, self-assessed skin phototypes Korea
are unreliable in Korean brown skin. Accepted for publication July 8, 1998

Many attempts have been made to predict individ-
ual sensitivity to ultraviolet (UV) radiation.
Among the variable methods, skin phototyping is
a simple and practical screening tool for estimating
the relative risk of exposure to UV light (1, 2); it
represents the skin’s response to sun exposure, ac-
cording to the history of a person’s sun-burning
and sun-tanning experience (1). Kang et al. (3) re-
ported an 85% agreement between skin phototypes
reported with a self-administered test and those
determined when the same patients were interview-
ed by a physician, and that these phototypes dem-
onstrated a positive correlation with the minimal
erythema dose (MED). This concept is, however,
based on skin responses in Caucasians. Several in-
vestigators have suggested that skin phototyping
does not reliably predict the UV response of indi-
viduals with pigmented skin (4, 5), nor even that
of Caucasians (6). Chung et al. (7) reported that
phototype did not correlate positively with MEDs,
suggesting that for those with brown skin, photo-
type is not useful. In addition, we previously re-
ported a weak relationship, if any, between skin
phototype and MED values (8). On the other
hand, subjects with MEDs lower than 60 mJ/cm2
were more commonly of phototypes II and III
rather than IV and V.
The pitfall of skin phototyping is that the assess-
ment of burning and tanning tendencies is based
on the assumption that there is an inverse relation-
ship between them. In other words, subjects who
always tend to burn should never tan (skin photo-
type I), and vice versa for skin phototype VI.
Rampen et al. (6), however, reported little recipro-
city between the tendency to burn and tan, and
that in Caucasians the correlation between burning
and tanning was poor.
In this study, we employed a self-assessment in-
strument to classify burning and tanning tenden-
cies (components of skin phototype), allowing us
to determine in brown skin the relationships of
skin phototype and MED with the tendency to
burn and to tan.
Material and methods
Subjects
Seven hundred and thirty-three normal, healthy,
male, Korean medical college students, aged 23–
29 (average, 24.8) years, participated in this self-

160
 Self-assessment of skin phototype for Korean brown skin
assessed questionnaire study. In addition, in a sub-
group of 156 men the minimal erythema dose
(MED) was determined. No subject had a history
of drug hypersensitivity or photosensitivity, none
had taken drugs during the 2 weeks prior to the
study, and their backs had not been exposed to the
sun during the preceding 3 months.
Self-assessed questionnaires
Tendency to burn
Each subject was asked about their tendency to
burn – the response of their skin 1 day after 1 h of
sun exposure during early summer – according to
the following categories: B1, always burns easily;
B2, burns moderately; B3, rarely burns.
Tendency to tan
Subjects were asked about their tendency to tan –
the response of their skin 1 week after regular sun
exposure during summer – according to the follow-
ing categories: T1, never tans; T2, tans minimally
with difficulty; T3, tans gradually and uniformly
(light brown); T4, tans profusely (moderate/dark
brown).
Skin phototype catergorized by the combination of
self-assessed tendency to burn and tan
Skin phototypes were determined by the combi-
nation of self-assessed tendencies to burn and tan.
The combination of burning tendency B1 and tan-
ning tendency T1 indicated skin phototype I; B1
and T2, II; B2 and T3, III; B2 and T4, IV; B3 and
T4, V. Skin phototype VI was omitted.
an F75/85W/UV21 fluorescent lamp that emitted
UV between 285 and 350 (peak, 310–315) nm. En-
ergy was measured with a Waldmann UV meter
(Waldmann Co., Germany). MED was defined as
the smallest exposure dose needed to produce
minimally perceptible erythema 24 h after ir-
radiation. With the light source mentioned above,
1¿1 cm2 skin sites were irradiated by 10 mJ/cm2
stepwise increments starting from 20 mJ/cm2 of
UVB. The MED values were stratified according
to the tendency to burn and to tan.
Data analysis
To normalize the data and stabilize the variance,
tests were performed on natural logarithmic trans-
formations of MEDs. These mean logarithmic
values for burning and tanning tendencies were
compared by analysis of variance (ANOVA). A P-
value ∞0.05 indicated statistical significance.
Results
Tendency to burn (Fig. 1)
The results of burning tendency were as follows:
Of 733 subjects, only 25 (3.4%) stated that they
always burned easily. Four hundred and four
(55.1%) stated that they burned minimally and 304
(41.5%) that they rarely burned.
Tendency to tan (Fig. 1)
Of 733 subjects, 50 (6.8%) stated that they never
tanned; 272 (37.1%) tanned minimally with diffi-

Self-assessed skin phototype

Finally, after listening to our explanation of the
definition of Fitzpatrick skin phototypes, subjects
were asked to identify their own type; this was de-
termined using the following criteria, as proposed
by Fitzpatrick (2). The definition consisted of two
phrases, which represented burning and tanning
tendencies, respectively.
I, always burns easily; never tans.
II, always burns easily; tans minimally with dif-
ficulty.
III, burns minimally; tans gradually and uni-
formly (light brown).
IV, burns minimally; tans profusely (moderate
brown).
V, rarely burns; tans profusely (dark brown).

Determination of minimal erythema dose

For a subgroup of 156 students, MED values were
determined. The UVB light source, a Waldmann Fig. 1. Proportions of subjects according to the tendency to
UV-800 (Waldmann Co., Germany), consisted of burn (1) and tan (2).

161
Park et al.
3 & 4), neither a significant difference nor a reliable
pattern was noted. If, however, the small numbers
of those who never tanned were ignored (T1: nΩ9,
1.4%), groups with a greater tendency to tan
showed higher MEDs. No significant differences
were found between tanning tendency T2 and T3
(PΩ0.9321), but each was significantly lower than
the mean MED of tanning tendency T4 (P∞0.05).

Fig. 2. Distribution of subjects accroding to history of burning
and tanning tendencies. The thick outline indicates Fitzpa-
trick’s skin phototype classified according to the combination
of tendency to burn and tan. The numbers in parentheses indi-
cate cases in which there is agreement between skin phototype
chosen by themselves and classified from the combination of
tendency to burn and tan (total numberΩ733). [B1, always
burns easily; B2, burns moderately; B3, rarely burns. T1, never
tans; T2, tans minimally with difficulty; T3, tans gradually and
uniformly (light brown); T4, tans profusely (moderate/dark
brown).]
Discussion
Since Fitzpatrick (9) proposed the concept of skin
response to UV exposure, skin phototype has been
used widely for clinical purposes. The method is
simple and practical, though due to concerns
about objectivity, reliability and reproducibility, its
validity has been questioned. We, too, have sug-
gested that there is a weak relationship between

culty; 285 (38.9%) tanned gradually and uniformly

(light brown); and 126 (17.2%) tanned profusely
(moderate/dark brown).

Skin phototype categorized by combining self-assessed

tendencies to burn and tan
Only 322 of 733 cases (43.9%) were classifiable ac-
cording to established criteria (Fig. 2); 411 did not
fit any skin phototype category, indicating a poor
relationship between burning and tanning tenden- Fig. 3. Relationship between self-assessed burning tendency and
cies. Skin phototype I was found in six people, and minimal erythema dose (MED) values. No significant differ-
type II in eight; phototype III was predominant, ence was found between each group (P±0.05). (B1, always
accounting for 187 cases, while phototypes IV and burns easily; B2, burns moderately; B3, rarely burns.)
V accounted for 47 and 74 cases, respectively.

Self-assessed skin phototype

Each subject was asked to decide his or her own
skin phototype (Fig. 2), but because this did not
clearly correspond to any one category, many hesi-
tated (data not shown). In the matching of self-
assessed and categorized skin phototypes, both
classified according to a combination of burning
and tanning tendencies, there was agreement in a
markedly low number of cases (252/733, 34.4%).

Determination of minimal erythema dose

In a subgroup in which MED had been deter-
mined, 93 of 156 cases (59.6%) were classified
strictly according to category. In 81 of 156 cases
(51.9%), the classified and self-assessed phototypes
matched. When MED values were compared with
self-assessed burning and tanning tendency (Figs.
Fig. 4. Relationship between self-assessed tanning tendency and
minimal erythema dose (MED) values. MEDs of tanning ten-
dency T2 and T3 were significantly lower than that of tanning
tendency T4 (P∞0.05), but no significant difference was found
among other groups (*P∞0.05 vs tanning tendency T4). [T1,
never tans; T2, tans minimally with difficulty; T3, tans gradually
and uniformly (light brown); T4, tans profusely (moderate/dark
brown).]

162
 Self-assessment of skin phototype for Korean brown skin
skin phototype and MEDs, especially in brown
skin (8).
The basic concept of skin phototype is that there
is a reciprocal relationship between the tendency
to burn and the tendency to tan. In other words,
a person who always burns should never tan (skin
phototype I), and one who always tans should
never burn (skin phototype VI). Between these two
extremes there are other subgroups, classified ac-
cording to combinations of these tendencies. On
the other hand, Rampen et al., in a study involving
Caucasians, reported that only 41.1% of cases were
classifiable according to the skin phototype, and
that there was little reciprocal interdependence be-
tween the tendency to burn and tendency to tan
(6).
In our study, over half of the subjects (56.1%)
did not show a correlation between burning and
tanning tendencies. Even though they were medical
students and we explained the definition of each
skin phototype, many subjects (data not shown)
hesitated in choosing their own skin phototype, on
the grounds that their characteristics did not corre-
spond to any one type. In addition, the concordant
ratio between skin phototype classified by the com-
bination of burning and tanning tendencies and
self-selected skin phototype was much lower.
In a subgroup in which MEDs were determined,
no significant correlation was found between MED
values and burning tendency, which was contrary
to our expectation. Similar results were observed
in the study of Rampel et al., in which there was
a poor correlation between burning tendency and
MEDs in Caucasians (6). These investigators dis-
cussed various sources of bias, such as limited
numbers of subjects with skin phototype I and II,
the use of combined burning and tanning propen-
sities and inadequate responses to self-reporting
questions. In fact, if the small number of ‘‘never
tan’’ cases in our data are ignored (T1; nΩ9, 1.4%),
higher MEDs indicate a greater propensity to tan.
Although no significant differences were found be-
tween tanning tendencies T2 and T3 (PΩ0.9321),
their MEDs were significantly lower than that of
subjects with tanning tendency T4 (P∞0.05). These
results agree with those of previous studies show-
ing that the tendency to tan correlates better with
skin complexion characteristics than does the ten-
dency to burn (1, 6), and suggesting that the ten-
dency to tan may be a more important predictor
of individual sensitivity to UV exposure, especially
in brown skin.
In conclusion, it appears that burning and tan-
ning tendencies, components of skin phototype,
are not inversely correlated. In Korean brown skin,
when comparing MED values with tendencies to
burn and tan, tanning tendency correlates better
than burning tendency with MEDs.
Acknowledgments
This study was supported by a grant from the Korean Ministry
of Health and Welfare as a part of a project of Good Health R
and D project ’97.
References
1. Fitzpatrick TB. The validity and practicability of sun-reac-
tive skin types I through VI. Arch Dermatol 1988: 124:
869–871.
2. Pathak MA, Fitzpatrick TB. Preventive treatment of sun-
burn, dermatoheliosis, and skin cancer with sun-protective
agent. In: Fitzpatrick TB, Eisen AZ, Wolff K et al., eds.
Dermatology in general medicine, 4th edn. New York:
McGraw-Hill, 1993: 1689–1696.
3. Kang S, Fitzpatrick TB, Youn JI et al. Identification of
population-at-risk for melanoma and epithelial skin cancer
using skin phototyping. In: Book of Abstracts of 18th
World Congress of Dermatology. 1992: 109A.
4. Leenutaphong V. Relationship between skin color and cu-
taneous response to ultraviolet radiation in Thai. Photo-
dermatol Photoimmunol Photomed 1995: 11: 198–203.
5. Stanford DG, Georgouras KE, Sullivan EA, Greenoak GE.
Skin phototyping in Asian Australians. Australas J Derma-
tol 1996: 37: suppl (1): s36–s38.
6. Rampen FHJ, Fleuren BAM, de Boo TM, Lemmens
WAJG. Unreliability of self-reported burning tendency and
tanning ability. Arch Dermatol 1988: 124: 885–888.
7. Chung JH, Koh WS, Youn JI. Relevance of skin phototyp-
ing to a Korean population. Clin Exp Dermatol 1994: 19:
476–478.
8. Youn JI, Oh JK, Kim BK et al. Relationship between skin
phototype and MED in Korean, brown skin. Photoderm-
atol Photoimmunol Photomed 1997: 13: 208–211.
9. Fitzpatrick TB. Soleil et peau. J Med Esthet 1975: 2: 33–
34.
163