Personal Use of Hair Dyes and Risk of Cancer

Abstract

Use of hair dyes has been suggested recently as a risk factor for several types of cancer in epidemiologic studies. This alarming news and controversial declarations by scientific organizations and general media have made necessary a systematic evaluation of the epidemiologic evidence.
To examine the association between personal use of hair dyes and relative risk of cancer.
We retrieved studies published in any language by systematically searching the MEDLINE (1966-January 2005), EMBASE, LILACS, and ISI Proceedings computerized databases and by manually examining the references of the original articles, reviews, and monographs retrieved.
We included cohort and case-control studies reporting relative risk estimates and 95% confidence intervals (CIs) (or data to calculate them) of personal hair dye use and cancer. We excluded studies that dealt with occupational exposure. We carried out separate analyses for bladder, breast, and hematopoietic cancers and cancers of other sites. Seventy-nine studies were included of 210 articles identified in the search.
Data were extracted independently by 2 investigators. We used a standardized questionnaire to record information on study design, sample size, type of controls, year of publication, adjustment factors, and relative risks of cancer among ever users of hair dyes. When possible, we extracted association measures on use of permanent dyes and extensive use (>200 lifetime episodes of dye use).
Study-specific relative risks were weighted by the inverse of their variance to obtain fixed- and random-effects pooled estimates. The pooled relative risk for ever users of hair dyes was 1.06 (95% CI, 0.95-1.18) for breast cancer (14 studies), 1.01 (95% CI, 0.89-1.14) for bladder cancer (10 studies), and 1.15 (95% CI, 1.05-1.27) for hematopoietic cancers (40 studies). Other cancers were examined by only 1 or 2 studies, of which the pooled or single relative risk was elevated for brain cancer, ovarian cancer, and cancer of the salivary glands. No effect was observed for use of permanent dyes or for extensive use.
We did not find strong evidence of a marked increase in the risk of cancer among personal hair dye users. Some aspects related to hematopoietic cancer and other cancers that have shown evidence of increased risk in 1 or 2 studies should be investigated further.

Full text

REVIEW
Personal Use of Hair Dyes and Risk of Cancer
A Meta-analysis
Bahi Takkouche, MD, PhD
Mahyar Etminan, PharmD, MSc
Agustín Montes-Martínez, MD, PhD
T
HERE IS GROWING CONCERN
worldwide about a possible in-
crease in the risk of cancer
among users of hair dyes. Some
aromatic amines contained in hair dyes
are mutagenic in vitro
1
and are carcino-
genic in animals and humans.
2,3
Prompted by the publication of a study
that showed a positive relation be-
tween hair dyes and bladder cancer,
4
the
Scientific Committee on Cosmetic Prod-
ucts and Non-food Products (Euro-
pean Commission) called for an urgent
review of the information and con-
cluded in 2004 that “bladder cancer may
be caused by carcinogenic arylamines in
the hair dye solutions.”
5
The US Food
and Drug Administration declared re-
cently that “FDA continues to follow re-
search in this field,”
6
while the Interna-
tional Agency for Research on Cancer
has stated that there is “inadequate evi-
dence of carcinogenicity” of hair dyes.
7
In 2002, the British Broadcasting Cor-
poration (BBC) caused alarm by report-
ing that “the Cancer Research Society ad-
vised consumers not to use hair products
until further research has been com-
pleted.”
8
However, it was not possible
to confirm the authorship and veracity
of this declaration.
As indicated by these statements,
health organizations are concerned
about the carcinogenicity of hair dyes,
although their position sways be-
tween skeptical vigilance and official
calls for bans of these products.
An association between hair dyes and
cancer would be an important public
health concern since about one third of
women in Europe and North America,
along with 10% of men older than 40
years, use some type of hair dye.
9
Per-
manent dyes, the most aggressive type,
represent 70% of the market share—
even more in Asia.
7
Recent reviews have commented
on the association between dye use
and cancer.
7,9-11
However, no meta-
Author Affiliations: Department of Preventive Medi-
cine, University of Santiago de Compostela, San-
tiago de Compostela, Spain (Drs Takkouche and Mon-
tes-Martínez); Division of Clinical Epidemiology, Royal
Victoria Hospital, Montreal, Quebec (Dr Etminan); and
Center for Clinical Epidemiology and Evaluation, Van-
couver Coastal Health Institute, Vancouver, British Co-
lumbia (Dr Etminan).
Corresponding Author: Bahi Takkouche, MD, PhD,
Department of Preventive Medicine, Faculty of Medi-
cine, University of Santiago de Compostela, 15782 San-
tiago de Compostela, Spain (mrbahi@usc.es).
Context Use of hair dyes has been suggested recently as a risk factor for several types
of cancer in epidemiologic studies. This alarming news and controversial declarations
by scientific organizations and general media have made necessary a systematic evalu-
ation of the epidemiologic evidence.
Objective To examine the association between personal use of hair dyes and rela-
tive risk of cancer.
Data Sources We retrieved studies published in any language by systematically search-
ing the MEDLINE (1966–January 2005), EMBASE, LILACS, and ISI Proceedings com-
puterized databases and by manually examining the references of the original articles,
reviews, and monographs retrieved.
Study Selection We included cohort and case-control studies reporting relative risk
estimates and 95% confidence intervals (CIs) (or data to calculate them) of personal
hair dye use and cancer. We excluded studies that dealt with occupational exposure.
We carried out separate analyses for bladder, breast, and hematopoietic cancers and
cancers of other sites. Seventy-nine studies were included of 210 articles identified in
the search.
Data Extraction Data were extracted independently by 2 investigators. We used a
standardized questionnaire to record information on study design, sample size, type
of controls, year of publication, adjustment factors, and relative risks of cancer among
ever users of hair dyes. When possible, we extracted association measures on use of
permanent dyes and extensive use (⬎200 lifetime episodes of dye use).
Data Synthesis Study-specific relative risks were weighted by the inverse of their
variance to obtain fixed- and random-effects pooled estimates. The pooled relative
risk for ever users of hair dyes was 1.06 (95% CI, 0.95-1.18) for breast cancer (14
studies), 1.01 (95% CI, 0.89-1.14) for bladder cancer (10 studies), and 1.15 (95% CI,
1.05-1.27) for hematopoietic cancers (40 studies). Other cancers were examined by
only 1 or 2 studies, of which the pooled or single relative risk was elevated for brain
cancer, ovarian cancer, and cancer of the salivary glands. No effect was observed for
use of permanent dyes or for extensive use.
Conclusions We did not find strong evidence of a marked increase in the risk of
cancer among personal hair dye users. Some aspects related to hematopoietic cancer
and other cancers that have shown evidence of increased risk in 1 or 2 studies should
be investigated further.
JAMA. 2005;293:2516-2525 www.jama.com
2516 JAMA, May 25, 2005—Vol 293, No. 20 (Reprinted) ©2005 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ on 02/25/2013

analysis has been carried out so far. We
therefore summarized the scientific evi-
dence following the Meta-analysis of
Observational Studies in Epidemiol-
ogy (MOOSE) guidelines for meta-
analyses of observational studies.
12
METHODS
Systematic Search
We conducted a computerized
MEDLINE search from 1966 to Janu-
ary 2005 to identify potentially eligible
studies. A study was defined as an analy-
sis of a specific cancer-exposure asso-
ciation; thus, a single article or publica-
tion could report several studies. We
applied the following algorithm both in
Medical Subject Heading and in free-
text words: (hair) and (dye* or colour*
or color*) and (cancer* or neoplasm*
or carcinogen*). To check whether ev-
ery article on the topic was retrieved, we
performed a second search, introduc-
ing the words cancer, neoplasms, and hair
dye in an unstructured fashion. We used
similar strategies to search EMBASE
(1980-2004) and LILACS (Latin
America and Caribbean) databases. We
searched meeting abstracts using the ISI
Proceedings database from its incep-
tion in 1990 to 2004. We also exam-
ined the references of every article re-
trieved and those of recent reviews and
monographs of hair dyes and can-
cer.
9-11
We considered including any rel-
evant article, independent of the lan-
Table 1. Study-Specific RRs of Female Breast Cancer and Hair Dye Use
Source
RR (95% CI)
Type of
Controls
Adjustment, Matching, and
Restriction Factors
Cases/Controls
or Cohort SizeHair Dye (Any) Permanent Dye
Intensive
Exposure*
Case-Control Studies
Kinlen et al,
21
1977 1.01 (0.71-1.44) . . . 0.69 (019-2.25) H Age, marital status, social
class
191/561
Shore et al,
22
1979 0.98 (0.75-1.27) 1.04 (0.79-1.37) 1.28 (0.58-2.79) H Age, duration of hair dye use 129/193
Nasca et al,
23
1980 1.28 (0.80-2.05) . . . 1.70 (0.52-5.52) P Age, county of residence,
unspecified others
118/233
Stavraky et al,
24,25
1979, 1981
(Toronto)
1.1 (0.5-2.7) 1.1 (0.5-2.4) . . . P Age, smoking, family history
of cancer, age at first
birth
35/70
Stavraky et al,
24,25
1979, 1981
(London)
1.2 (0.6-2.6) 1.3 (0.6-2.5) . . . H Age, smoking, family history
of cancer, age at first
birth
50/100
Wynder and
Goodman,
26
1983
1.02 (0.78-1.32) . . . 1.18 (0.84-1.64) H Age, religion, education,
marital status, smoking
401/625
Koenig et al,
27
1991 0.8 (0.6-1.1) 0.85 (0.70-1.03) 0.8 (0.6-1.2) H Age, family history of cancer,
age at first birth,
birthplace, race, religion,
history of receiving
Medicaid
398/790
Nasca et al,
28
1992 1.04 (0.90-1.21) 1.00 (0.86-1.17) 1.01 (0.82-1.24) P Age, county of residence 1617/1617
Boice et al,
29
1995 1.08 (0.87-1.33) . . . . . . P Age, age at menarche,
menopause, and first
birth, family history of
breast cancer
528/2628
Cook et al,
30
1999 1.3 (1.0-1.6) 1.00 (0.7-1.3) . . . P Age, parity, weight, income,
education, marital status,
family history of breast
cancer, smoking, alcohol
consumption
844/960
Zheng et al,
31
2002 0.9 (0.7-1.2) 0.9 (0.7-1.2) 1.9 (0.9-4.0) P Age, race, menopause, study
site, history of breast
cancer, history of
lactation, fat intake
608/609
Petro-Nustas et al,
32
2002
8.62 (3.33-
22.28)
. . . . . . P Age, parity, education, place
of residence
100/100
Cohort Studies
Green et al,
33
1987 . . . 1.1 (0.9-1.2) 0.98 (0.71-1.37) . . . Age, age at first birth,
smoking, history of breast
cancer, menopausal
status, history of benign
breast disease
353/118 404
Altekruse et al,
34
1999
. . . 0.9 (0.9-1.0) 0.9 (0.8-1.1) . . . Age, race, duration of hair
dye use, smoking
782/547 589†
Abbreviations: CI, confidence interval; H, hospital; P, population; RR, relative risk. Ellipses indicate data not applicable.
*Defined as more than 200 lifetime exposures to hair dye.
†Based on fatal cancer cases.
PERSONAL USE OF HAIR DYES AND RISK OF CANCER
©2005 American Medical Association. All rights reserved. (Reprinted) JAMA, May 25, 2005—Vol 293, No. 20 2517
Downloaded From: http://jama.jamanetwork.com/ on 02/25/2013

guage of the publication. Unpublished
studies were not considered. Formal
translation was necessary in only 1 ar-
ticle published in Japanese.
13
All searches
were carried out independently by 2 epi-
demiologists (B.T. and A.M.-M.) and re-
sults were merged.
Inclusion Criteria
and Data Collection
Studies were included if (1) they pre-
sented original data from case-control
or cohort studies; (2) the outcome of
interest was clearly defined as cancer
of an anatomical site; (3) the exposure
of interest was personal hair dye use;
and (4) they provided relative risk (RR)
estimates and their confidence inter-
vals (CIs) or provided enough data to
calculate them (raw data, P value, or
variance estimate). We did not con-
sider studies that dealt with occupa-
tional exposure to hair dyes or those
that concerned childhood cancers re-
lated to use of hair dye by the mother.
If data were duplicated in more than 1
study, the most recent study was in-
cluded in the analysis. We contacted the
authors of the publications when fur-
ther explanations were necessary to as-
sess whether the information pro-
vided in more than 1 article concerned
the same study. When RRs of cancer of
different anatomical sites were avail-
able in the same publication, we con-
sidered each cancer separately.
We developed a questionnaire and
recorded study name, year of publica-
tion, study design, sample size (cases
and controls or cohort size), type of
controls for case-control studies (pa-
tients with other diseases or population-
based controls), variables used for ad-
justment or matching, and association
measures that compared ever users of
hair dyes with never users for any (or
unspecified) type of dye. When avail-
able, we also extracted association mea-
sures and CIs corresponding to the ex-
clusive use of permanent dye and
carried out a separate analysis for this
type of dye. If a study provided infor-
mation on permanent dye only and not
on use of hair dye of any kind, we in-
cluded this information in the pooled
analysis of hair dye of any type. To
check for possible changes in the re-
sult, we later recalculated our pooled
estimate for hair dye of any type ex-
cluding studies that provided informa-
tion on permanent dye only. Results
Table 2. Pooled RRs of Female Breast Cancer and Hair Dye Use
No. of
Studies
Fixed-Effects
RR (95% CI)
Random-Effects
RR (95% CI) R
i
*
P Value
(by Q Test)
All studies (any dye use) 14 1.04 (0.98-1.09) 1.06 (0.95-1.18) 0.68 ⬍.001
Cohort studies 2 1.01 (0.95-1.08) 1.00 (0.82-1.21) 0.88 .002
Case-control studies 12 1.07 (0.98-1.15) 1.09 (0.94-1.25) 0.62 .003
Population-based 7 1.12 (1.01-1.23) 1.21 (0.96-1.52) 0.77 .001
Hospital-based 5 0.96 (0.84-1.11) 0.96 (0.84-1.11) 0.00 .76
Permanent dye use only 9 1.00 (0.94-1.05) 0.98 (0.91-1.07) 0.40 .13
Intensive exposure† 9 0.99 (0.89-1.11) 0.99 (0.89-1.11) 0.00 .45
Abbreviations: CI, confidence interval; RR, relative risk.
*Proportion of total variance due to between-study variance.
†Defined as more than 200 lifetime exposures to hair dye.
Table 3. Study-Specific RRs of Bladder Cancer and Hair Dye Use
Source
RR (95% CI)
Type of
Controls
Adjustment, Matching,
and Restriction Factors
Cases/Controls
or Cohort SizeHair Dye (Any)
Permanent
Dye
Intensive
Exposure*
Case-Control Studies
Jain et al,
35
1977 1.1 (0.41-3.03) . . . . . . H Age, sex 107/107
Neutel et al,
36
1978
0.92 (0.38-2.24) . . . . . . H Age, sex 50/50
Howe et al,
37
1980
0.7 (0.3-1.4) . . . . . . P Age, sex 632/632
Stavraky et al,
25
1981
1.1 (0.4-2.8) . . . . . . P Sex, unspecified others 23/46
Hartge et al,
14
1982
1.0 (0.9-1.2) . . . 0.82 (0.65-1.03) P Age, sex, race, smoking 2982/5782
Ohno et al,
38
1985 1.49 (0.89-2.5) . . . . . . P Age, sex, smoking 293/589
Nomura et al,
39
1989
1.41 (0.86-2.33) . . . . . . P Age, sex, race, place of
residence, smoking
261/522
Gago-Dominguez
et al,
4
2001
1.0 (0.7-1.4) 1.3 (0.8-2.0) 1.83 (1.18-2.84) P Age, sex, race, place of
residence, smoking
897/897
Andrew et al,
40
2004
0.70 (0.49-1.00) 1.15 (0.70-1.88) 1.9 (0.7-4.8) P Age, sex, smoking,
education level
495/665
Cohort Study
Henley and
Thun,
41
2001
. . . 1.08 (0.84-1.38) . . . . . . Age, sex, race, smoking,
education,
occupation
336/547 571†
Abbreviations: CI, confidence interval; H, hospital; P, population; RR, relative risk. Ellipses indicate data not applicable.
*Defined as more than 200 lifetime exposures to hair dye.
†Based on fatal cancer cases.
PERSONAL USE OF HAIR DYES AND RISK OF CANCER
2518 JAMA, May 25, 2005—Vol 293, No. 20 (Reprinted) ©2005 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ on 02/25/2013

were similar in both analyses, so only
the former was presented in the tables.
When possible, we also extracted RRs
for intensive use of hair dye, defined as
200 or more lifetime episodes of dye-
ing. If not directly available in the pub-
lications, we calculated this magni-
tude assuming an average frequency of
hair dyeing of 11.7 times per year
among women. This number was ob-
tained from the distribution of hair dye
use among the control group of 1 of the
studies included in our meta-analysis,
which provided detailed information on
frequency of use.
14
Quality Assessment
We assessed study quality based on a
10-point scale that included elements
of previous published scales adapted to
the needs of the present meta-
analysis.
15
Each study was scored ac-
cording to 5 characteristics of meth-
ods and presentation of results. Each
item was scored from 0 to 2. Specifi-
cally, for case-control studies, we de-
termined whether participation rate was
at least 80% in both groups; whether
cases were incident, prevalent, or dead;
whether controls were taken from 1 or
various hospitals or from the general
population; whether potential con-
founding for sex, age, and smoking was
corrected or prevented through match-
ing or adjustment; and whether dura-
tion of exposure to hair dye was accu-
rately measured. For cohort studies, in
addition to the criteria listed above that
were not specific to case-control de-
signs, we determined whether loss to
follow-up was less than 20% of the ini-
tial cohort size and whether efforts were
made to ensure that the cohort did not
change exposure during follow-up.
(The complete protocol for quality scor-
ing is available on request from the au-
thors.) For stratification purposes, stud-
ies that scored 7 or higher of 10 were
considered to be of high quality and the
rest to be of low quality. Quality scor-
ing was performed independently by 2
reviewers (B.T. and A.M.-M.) and the
average score between reviewers was as-
signed to the studies. Disagreement was
measured by the Bland-Altman limits-
of-agreement method.
16
The average
disagreement was close to zero (−0.22;
95% CI, −2.42 to 1.98). The fact that
the 95% CI is symmetric around zero
suggests that there was no systematic
disagreement.
Statistical Analysis
We weighted the study-specific ad-
justed log odds ratios for case-control
studies and log RRs for cohort studies
by the inverse of their variance to com-
pute a pooled RR and its 95% CI. We
presented both fixed- and random-
effects pooled estimates but preferen-
tially used the latter when heterogene-
ity was present.
We used a parametric bootstrap ver-
sion (1000 replications) of the DerSi-
monian and Laird Q test to check for
heterogeneity.
17
The null hypothesis of
this test is absence of heterogeneity. To
quantify this heterogeneity, we calcu-
lated the proportion of the total vari-
ance due to between-study variance (R
i
statistic).
17
To further explore the ori-
gin of heterogeneity, we restricted the
analysis to subgroups of studies de-
fined by study characteristics such as
case-control/cohort design, adjust-
ment factors, and type of controls (hos-
pital-based or population-based).
We used funnel plots to assess pub-
lication bias visually. Because funnel
plots have several limitations and rep-
resent only an informal way to detect
publication bias,
18
we carried out more
formal testing using the test proposed
by Egger et al.
19
All analyses were per-
formed with the software HEpiMA, ver-
sion 2.1.3,
20
and STATA, version 8.0
(Stata Corp, College Station, Tex).
RESULTS
We identified 79 studies, carried out in
11 countries, on personal hair dye use
and cancer that met our inclusion cri-
teria and were included in the analysis.
We found 14 studies on breast can-
cer,
21-34
10 studies on bladder can-
cer,
4,14,25,35-41
and 40 studies on hemato-
poietic cancers.
25,34,42-58
Furthermore, 2
studies provided data on adult brain tu-
mors,
59,60
2 on skin cancer,
61-63
2 on ovar-
ian cancer,
25,64
and 2 on cervical can-
cer.
25,51
We also found 1 study for each
of the following cancer sites: salivary
gland,
65
endometrium,
25
vagina,
51
oral
cavity,
51
soft tissue sarcoma,
58
digestive
system,
51
and respiratory system.
51
Among the studies that could have
been relevant to our meta-analysis, 1
was excluded because it presented
cross-sectional data
66
and another be-
cause it did not present any measure of
uncertainty of the RR.
67
Breast Cancer
The 12 case-control studies (involv-
ing 5019 cases and 8486 controls) and
2 cohort studies that dealt with breast
cancer were published between 1977
and 2002 (T
ABLE 1). Seven case-
control studies used population-
based controls and 5 used hospital-
based controls.
Compared with never use of hair
dyes, the random-effects pooled RR of
breast cancer for any type of dye use was
1.06 (95% CI, 0.95-1.18). The pooled
RR for exclusive use of permanent dye
was 1.00 (95% CI, 0.94-1.05) and was
0.99 (95% CI, 0.89 to 1.11) for inten-
sive exposure (T
ABLE 2). There was no
substantial difference in pooled RRs
Table 4. Pooled RRs of Bladder Cancer and Hair Dye Use
No. of
Studies
Fixed-Effects
RR (95% CI)
Random-
Effects
RR (95% CI) R
i
*
P Value
(by Q Test)
All studies (any dye use) 10 1.01 (0.89-1.14) 1.01 (0.89-1.14) 0.05 .41
Case-control studies 9 0.99 (0.87-1.13) 0.99 (0.85-1.15) 0.14 .35
Population-based 7 0.99 (0.87-1.13) 1.00 (0.83-1.20) 0.38 .15
Hospital-based 2 0.82 (0.47-1.43) 0.82 (0.47-1.43) 0.00 .42
Permanent dye use only 3 1.13 (0.93-1.38) 1.13 (0.93-1.38) 0.00 .78
Intensive exposure† 3 1.00 (0.82-1.22) 1.33 (0.69-2.56) 0.90 .001
Abbreviations: CI, confidence interval; RR, relative risk.
*Proportion of total variance due to between-study variance.
†Defined as more than 200 lifetime exposures to hair dye.
PERSONAL USE OF HAIR DYES AND RISK OF CANCER
©2005 American Medical Association. All rights reserved. (Reprinted) JAMA, May 25, 2005—Vol 293, No. 20 2519
Downloaded From: http://jama.jamanetwork.com/ on 02/25/2013

across designs (cohort, hospital-based
case-control, and population-based
case-control studies). Restricting the
analysis to the 7 studies that scored 7
points or higher on the quality scale did
not alter the results (RR, 1.06; 95% CI,
0.93-1.20). Heterogeneity of the study-
specific RRs was moderate to large for
case-control studies (R
i
=0.62) and all
studies analyzed together (R
i
=0.68).
This was explained mainly by the re-
sult of 1 population-based case-
control study
32
with an RR of 8.62 and
a wide 95% CI, which might be con-
sidered as an influential point. Exclud-
Table 5. Study-Specific RRs of Hematopoietic Cancers and Hair Dye Use
Source
RR (95% CI)
Type of
Controls
Adjustment, Matching,
and Restriction Factors
Cases/Controls
or Cohort SizeHair Dye (Any)
Permanent
Dye
Intensive
Exposure*
Non-Hodgkin Lymphoma
Case-control studies
Stavraky et al,
25
1981†
Toronto 0.7 (0.3-1.6) . . . . . . P Sex, unspecified others 45/90
London 1.2 (0.4-3.8) . . . . . . H Sex, unspecified others 25/50
Cantor et al,
42
1988‡ 2.0 (1.3-3.0) . . . . . . P Age, sex, place of residence,
smoking, pesticide exposure,
family history of cancer
53/1245
Zahm et al,
43
1992‡ 1.29 (0.93-1.81) 1.7 (1.1-2.8) 1.2 (0.6-2.7) P Age, sex, race, vital status 385/1432
Linos et al,
44
1994 2.07 (0.79-5.41) . . . 2.07 (0.79-5.41) H Age, sex 179/274
Holly et al,
45
1998 1.07 (0.86-1.34) 0.85 (0.61-1.19) 1.0 (0.7-1.6) P Age, sex, sexual preference,
number of sexual partners,
smoking, education
1593/2515
Miligi et al,
46
1999‡ 1.0 (0.8-1.2) . . . . . . P Age, sex, education, smoking 611/828
Schroeder et al,
47
2002 2.0. (1.3-2.9) . . . . . . P Age, sex, residency, vital status 182/1245
Zhang et al,
48
2004 1.1 (0.9-1.5) 1.2 (0.9-1.5) 1.18 (0.85-1.63) P Age, sex, family history of
cancer, race, education,
smoking, alcohol
consumption
449/519
Chiu et al,
49
2004
Iowa 1.99 (1.29-3.06) . . . . . . P Age, sex, race, place of
residence, vital status
605/1206
Nebraska 0.90 (0.47-1.71) . . . . . . P Age, sex, race, place of
residence, vital status
200/712
Tavani et al,
58
2005‡ 1.03 (0.73-1.44) 1.25 (0.87-1.79) . . . H Age, sex, place of residence,
smoking, education
446/1295
Cohort studies
Grodstein et al,
50
1994 . . . 1.1 (0.8-1.6) 0.9 (0.5-1.7) . . . Age, sex, smoking 144/99 067
Altekruse et al,
34
1999 . . . 1.1 (1.0-1.3) 1.0 (0.8-1.3) . . . Age, sex, race, duration of dye
use, smoking
350/573 369§
Hodgkin Disease
Case-control studies
Zahm et al,
43
1992‡ 1.7 (0.8-3.5) 3.2 (1.3-8.2) 5.0 (0.5-40.2) P Age, sex, race, vital status 70/1418
Tavani et al,
58
2005‡ 0.68 (0.40-1.18) 1.14 (0.63-2.08) . . . H Age, sex, place of residence,
smoking, education
158/1295
Cohort studies
Grodstein et al,
50
1994 . . . 0.9 (0.4-2.1) . . . . . . Age, sex, smoking 24/99 067
Thun et al,
51
1994 0.55 (0.23-1.36) . . . . . . . . . Age, sex, race, smoking 31/573 369§
Multiple Myeloma
Case-control studies
Zahm et al,
43
1992‡ 1.8 (1.0-3.3) 3.2 (1.3-7.9) 1.0 (0.1-5.0) P Age, sex, race, vital status 69/1418
Brown et al,
52
1992 1.9 (1.0-3.6) . . . . . . P Age, sex, vital status 173/650
Herrinton et al,
53
1994 1.14 (0.87-1.50) . . . . . . P Age, sex, race, study center,
education
689/1681
Tavani et al,
58
2005‡ 1.17 (0.70-1.97) 1.28 (0.75-2.19) . . . H Age, sex, place of residence,
smoking, education
141/1295
Cohort studies
Grodstein et al,
50
1994 . . . 0.4 (0.2-0.9) . . . . . . Age, sex, smoking 32/99 067
Altekruse et al,
34
1999 . . . 1.0 (0.8-1.3) 1.0 (0.7-1.5) . . . Age, sex, race, duration of dye
use, smoking
131/573 369§
(continued)
PERSONAL USE OF HAIR DYES AND RISK OF CANCER
2520 JAMA, May 25, 2005—Vol 293, No. 20 (Reprinted) ©2005 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ on 02/25/2013

ing this study resulted in low hetero-
geneity (R
i
=0.36). The funnel plot did
not provide evidence for publication
bias (data not shown). The Egger test
of asymmetry of the funnel plot yielded
a P value of .90, which did not change
substantially when the outlier was ex-
cluded (P=.72).
Bladder Cancer
The 10 studies, presented in TABLE 3,
that provided data on bladder cancer (9
case-control studies comprising 5740
cases and 9290 controls and 1 cohort
study with 336 cases) did not show sub-
stantial heterogeneity. The only excep-
tion was for intensive exposure, the het-
erogeneity of which was high but based
on only 3 studies.
The pooled RR (T
ABLE 4) did not
show any effect of hair dye on bladder
cancer (RR, 1.01; 95% CI, 0.89-1.14 for
all studies). This absence of effect was
constant across study designs, types of
exposure (permanent dyes and inten-
Table 5. Study-Specific RRs of Hematopoietic Cancers and Hair Dye Use (cont)
Source
RR (95% CI)
Type of
Controls
Adjustment, Matching,
and Restriction Factors
Cases/Controls
or Cohort SizeHair Dye (Any)
Permanent
Dye
Intensive
Exposure*
Leukemia
Case-control studies
Cantor et al,
42
1988‡
Acute nonlymphocytic 1.1 (0.5-2.6) . . . . . . P Age, sex, place of residence,
smoking, pesticide exposure,
family history of cancer
143/1245
Acute lymphocytic 2.9 (0.4-13.8) . . . . . . P Age, sex, place of residence,
smoking, pesticide exposure,
family history of cancer
16/1245
Chronic lymphocytic 1.4 (0.7-2.6) . . . . . . P Age, sex, place of residence,
smoking, pesticide exposure,
family history of cancer
243/1245
Chronic myeloid 2.2 (0.7-6.2) . . . . . . P Age, sex, place of residence,
smoking, pesticide exposure,
family history of cancer
51/1245
Other leukemia 3.3 (1.4-7.6) . . . . . . P Age, sex, place of residence,
smoking, pesticide exposure,
family history of cancer
61/1245
Zahm et al,
43
1992 (chronic
lymphocytic)‡
1.0 (0.5-2.2) 0.8 (0.1-4.0) . . . P Age, sex, race, vital status 56/1418
Mele et al,
54
1994‡
Acute myeloid 1.01 (0.78-1.30) . . . . . . H Age, sex, education, place of
residence
252/1161
Acute lymphocytic 1.2 (0.8-1.8) . . . . . . H Age, sex, education, place of
residence
100/1161
Chronic myeloid 1.1 (0.75-1.60) . . . . . . H Age, sex, education, place of
residence
156/1161
Markovic-Denic et al,
55
1995 (chronic
lymphocytic)
1.97 (1.08-3.59) . . . . . . H Age, sex, place of residence 130/130
Miligi et al,
46
1999 (all
leukemia)‡
0.9 (0.7-1.3) . . . . . . P Age, sex, education, smoking 260/828
Bjork et al,
56
2001 (chronic
myeloid)
0.35 (0.18-0.68) . . . . . . P Age, sex, place of residence 255/765
Rauscher et al,
57
2004
(acute)
1.3 (1.0-1.8) 1.6 (1.1-2.4) 1.00 (0.5-2.0) P Age, sex, race, education, place
of residence
769/623
Cohort studies
Grodstein et al,
50
1994
Chronic lymphocytic . . . 0.6 (0.3-1.5) . . . . . . Age, sex, smoking 23/99 067
Acute and chronic
myeloid and acute
lymphocytic
. . . 0.8 (0.3-1.9) . . . . . . Age, sex, smoking 21/99 067
Altekruse et al,
34
1999
(all leukemia)
. . . 1.1 (0.9-1.3) 1.3 (1.0-1.7) . . . Age, sex, race, duration of dye
use, smoking
718/573 369§
Abbreviations: CI, confidence interval; H, hospital; P, population; RR, relative risk. Ellipses indicate data not applicable.
*Defined as more than 200 lifetime exposures to hair dye.
†Cases are reported as “lymphoma and leukemia” cases in the original study.
‡These studies used the same control group for several outcomes.
§Based on fatal cancer cases.
PERSONAL USE OF HAIR DYES AND RISK OF CANCER
©2005 American Medical Association. All rights reserved. (Reprinted) JAMA, May 25, 2005—Vol 293, No. 20 2521
Downloaded From: http://jama.jamanetwork.com/ on 02/25/2013

sive exposure), and sex (RR, 1.03; 95%
CI, 0.90-1.17 for women and RR, 0.93;
95% CI, 0.77-1.13 for men). Since
smoking is an important risk factor for
bladder cancer, we recalculated our
pooled estimates for studies that pro-
vided data explicitly adjusted for to-
bacco consumption.
4,14,38-41
The pooled
RR was 1.05 (95% CI, 0.93-1.19).
Stratifying the analysis by quality
scoring did not show any difference in
the pooled RRs between studies with
good quality scores and the remain-
der. We could not detect any evidence
of publication bias on the funnel plot
(P=.75 by Egger symmetry test).
Hematopoietic Cancers
TABLE 5 and TABLE 6 show the specific
RRs for the 40 studies that dealt with he-
matopoietic cancers and the pooled RRs
for different types of these cancers. When
all hematopoietic cancers were ana-
lyzed together, we observed a border-
line increase in the risk (RR, 1.15; 95%
CI, 1.05-1.27) for ever users of hair dye.
This increase is restricted to case-
control studies (RR, 1.23; 95% CI, 1.09-
1.39) and, more specifically, to the 17
case-control studies with data on men
(RR, 1.57; 95% CI, 1.33-1.84). No co-
hort study with data on men was avail-
able for comparison. In these catego-
ries, heterogeneity was large. On the
contrary, we did not observe any risk in-
crease when we restricted our analysis
to women (RR, 1.04; 95% CI, 0.97-
1.11), to exclusive use of permanent dyes
(random-effects RR, 1.14; 95% CI, 0.99-
1.29), or to intensive exposure to hair
dyes. Stratification by quality scoring or
by adjustment for smoking did not pro-
Table 6. Pooled RRs of Hematopoietic Cancers and Hair Dye Use
No. of
Studies
Fixed-Effects
RR (95% CI)
Random-Effects
RR (95% CI) R
i
*
P Value
(by Q Test)
Non-Hodgkin lymphoma
All studies (any dye use) 14 1.16 (1.07-1.26) 1.23 (1.07-1.42) 0.57 .008
Cohort studies 2 1.10 (0.94-1.28) 1.10 (0.94-1.28) 0.00 .99
Case-control studies 12 1.19 (1.08-1.32) 1.27 (1.06-1.53) 0.63 .002
Population-based 9 1.20 (1.08-1.33) 1.29 (1.04-1.59) 0.72 ⬍.001
Hospital-based 3 1.31 (1.18-1.45) 1.31 (1.18-1.45) 0.00 .63
Permanent dye use only 6 1.13 (1.01-1.26) 1.13 (0.99-1.29) 0.19 .32
Intensive exposure† 6 1.07 (0.90-1.28) 1.07 (0.90-1.28) 0.00 .72
Hodgkin disease
All studies (any dye use) 4 0.87 (0.61-1.24) 0.88 (0.54-1.42) 0.42 .17
Permanent dye use only 3 1.34 (0.86-2.06) 1.41 (0.72-2.77) 0.57 .11
Multiple myeloma
All studies (any dye use) 6 1.11 (0.95-1.31) 1.14 (0.86-1.52) 0.62 .04
Cohort studies 2 0.92 (0.71-1.18) 0.69 (0.28-1.66) 0.91 .004
Case-control studies 4 1.28 (1.04-1.59) 1.31 (1.03-1.67) 0.17 .33
Permanent dye use only 4 1.04 (0.84-1.30) 1.10 (0.62-1.95) 0.83 .003
Leukemia
All studies (all leukemia types, any dye use) 16 1.10 (1.00-1.22) 1.12 (0.94-1.34) 0.54 .01
All leukemia types, studies with different sets of controls only 8 1.06 (0.94-1.21) 0.97 (0.74-1.28) 0.71 .01
Cohort studies (all leukemia types) 3 1.07 (0.91-1.25) 1.05 (0.87-1.28) 0.16 .34
Case-control studies (all leukemia types) 13 1.13 (0.99-1.28) 1.19 (0.95-1.49) 0.59 .004
Acute leukemia 5 1.14 (0.96-1.36) 1.14 (0.96-1.36) 0.00 .56
Chronic leukemia 6 1.11 (0.89-1.40) 1.13 (0.69-1.86) 0.76 .003
Myeloid leukemia 5 0.99 (0.82-1.18) 0.94 (0.64-1.39) 0.73 .02
Lymphocytic leukemia 5 1.41 (1.06-1.88) 1.41 (1.06-1.88) 0.00 .56
Permanent dye use only 5 1.13 (0.97-1.31) 1.12 (0.86-1.46) 0.48 .23
All hematopoietic cancers
All studies 40 1.13 (1.06-1.20) 1.15 (1.05-1.27) 0.52 ⬍.001
Studies with different sets of controls only 22 1.12 (1.04-1.21) 1.11 (0.97-1.28) 0.60 ⬍.001
Cohort studies 9 1.04 (0.94-1.15) 1.01 (0.89-1.16) 0.25 .27
Case-control studies 31 1.17 (1.09-1.26) 1.23 (1.09-1.39) 0.55 ⬍.001
Male-only studies 17 1.57 (1.33-1.84) 1.56 (1.32-1.85) 0.04 .43
Permanent dye use (all) 18 1.12 (1.04-1.22) 1.14 (0.99-1.29) 0.44 .03
Studies with different sets of controls 11 1.08 (0.99-1.18) 1.06 (0.94-1.20) 0.35 .15
Intensive exposure (all)† 11 1.12 (0.98-1.28) 1.12 (0.98-1.28) 0.00 .76
Studies with different sets of controls† 8 1.11 (0.97-1.27) 1.11 (0.97-1.27) 0.00 .69
Abbreviations: CI, confidence interval; RR, relative risk.
*Proportion of total variance due to between-study variance.
†Defined as more than 200 lifetime exposures to hair dye.
PERSONAL USE OF HAIR DYES AND RISK OF CANCER
2522 JAMA, May 25, 2005—Vol 293, No. 20 (Reprinted) ©2005 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ on 02/25/2013

duce any change in the results. The fun-
nel plot showed substantial asymmetry
(data not shown), confirmed by statis-
tical testing (P value of the inter-
cept=.02 by Egger test). Apparently,
there is a deficit of studies that show a
slightly protective but imprecise effect in
their results. Our calculations show that
only 5 such unpublished case-control
studies, with an RR of 0.8 and a 95% CI
of 0.6 to 1.1, would be sufficient to can-
cel the association between hair dye use
and risk of hematopoietic cancers and
render it nonsignificant (RR, 1.09; 95%
CI, 0.99-1.20).
Separate analyses by cancer site show
a slight increase in the risk of non-
Hodgkin lymphoma for ever users of
hair dye (RR, 1.23; 95% CI, 1.07-
1.42), which seems to be restricted to
case-control studies. However, the re-
sults of intensive exposure do not show
any association.
Case-control studies of multiple my-
eloma and lymphocytic leukemia show
a slight increase of the risk.
Other Cancers
TABLE 7 presents the study-specific RRs
of other cancer sites for hair dye users
vs never users. The single study on can-
cers of salivary glands shows a 3-fold
increase of the risk.
65
The pooled RRs
of the 2 studies available for each site
were 1.83 (95% CI, 1.16-2.89) for brain
tumors, 1.71 (95% CI, 1.15-2.53) for
ovarian cancer, 0.74 (95% CI, 0.51-
1.07) for skin cancer, and 0.89 (95% CI,
0.53-1.9) for cervical cancer.
COMMENT
Our results indicate that, globally, there
is no effect of personal hair dye use on
the risk of breast and bladder cancer.
There is a borderline effect for he-
matopoietic cancers. However, the evi-
dence of a causal effect is too weak to
represent a major public health con-
cern. The fact that the restriction of the
analysis to intensive exposure to hair
dyes and to exclusive use of perma-
Table 7. Study-Specific RRs of Other Cancers and Hair Dye Use
Source
Type of
Cancer Type of Study
RR (95% CI)
Adjusted, Matching,
or Restriction Factors
Cases/Controls
or Cohort SizeHair Dye (Any)
Permanent
Dye
Intensive
Exposure*
Ahlbom et al,
59
1986
Brain Case-control 1.5 (0.6-3.7) . . . . . . Age, sex, place of
residence
78/289
Burch et al,
60
1987 Brain Case-control 1.96 (1.15-3.32) . . . . . . Age, sex, place of
residence, marital
status, diagnosis
date
228/247
Holman and
Armstrong,
61,62
1983, 1985
Skin Case-control 2.40 (0.92-6.28) 1.1 (0.8-1.6) . . . Age, sex, birth place,
sunlight exposure
507/511
Osterlind et al,
63
1988
Skin Case-control 0.6 (0.5-0.9) 0.6 (0.4-1.0) . . . Age, sex 474/926
Stavraky et al,
25
1981
Ovary Case-control 0.91 (0.36-2.31) . . . . . . Unspecified 58/116
Tzonou et al,
64
1993
Ovary Case-control 1.96 (1.27-3.03) . . . Age, age at menarche,
at menopause, at
first birth,
education,
smoking, alcohol
consumption
189/200
Stavraky et al,
25
1981
Cervix Case-control 0.7 (0.3-1.9) . . . . . . Unspecified others 38/76
Thun et al,
51
1994 Cervix Cohort 0.97 (0.53-1.77) . . . 1.51 (0.63-3.59) Age, race, smoking Unspecified/573 369
Stavraky et al,
25
1981
Endometrium Case-control 1.6 (0.6-4.0) . . . . . . Unspecified 36/72
Spitz et al,
65
1990 Salivary gland Case-control 3.01 (1.18-7.69) . . . 3.5 (0.9-12.8) Age, sex, race 64/128
Thun et al,
51
1994 Vagina Cohort 0.93 (0.43-2.03) . . . 0.82 (0.19-3.47) Age, race, smoking Unspecified/573 369
Thun et al,
51
1994 Oral cavity Cohort 0.61 (0.32-1.14) . . . 0.31 (0.07-1.25) Age, sex, race,
smoking
Unspecified/573 369
Thun et al,
51
1994 Digestive
system
Cohort 0.94 (0.85-1.03) . . . 1.05 (0.90-1.23) Age, sex, race,
smoking
Unspecified/573 369
Thun et al,
51
1994 Respiratory
system
Cohort 1.00 (0.91-1.11) . . . 1.20 (1-02-1.42) Age, sex, race,
smoking
Unspecified/573 369
Tavani et al,
58
2005 Soft tissue
sarcoma
Case-control 0.73 (0.45-1.17) 1.23 (0.75-2.00) . . . Age, sex, place of
residence,
smoking,
education
221/1295
Abbreviations: CI, confidence interval; RR, relative risk.
*Defined as more than 200 lifetime exposures to hair dye.
PERSONAL USE OF HAIR DYES AND RISK OF CANCER
©2005 American Medical Association. All rights reserved. (Reprinted) JAMA, May 25, 2005—Vol 293, No. 20 2523
Downloaded From: http://jama.jamanetwork.com/ on 02/25/2013

nent hair dye does not strengthen the
risk further is consistent with the ab-
sence of causal effect. Furthermore,
there is room for publication bias as
shown by the funnel plot, the result of
the corresponding hypothesis test, and
the fact that only a small number of
negative studies, unpublished so far,
would cancel the association. None-
theless, the intriguing risk increase for
male users of hair dyes deserves fur-
ther consideration.
A plausible explanation for the lack
of effect may be the low concentration
of potential carcinogens in the hair dyes
and the fact that the use of some dye
ingredients, such as 2,4-diaminotolu-
ene and 2,4-diaminoanisole, was dis-
continued in the mid 1970s after they
were found to be carcinogenic in ro-
dents.
7
The overwhelming majority of
the studies were carried out several
years after this ban went into effect.
There is potential for recall bias from
case-control studies, which represent the
majority of the studies included in this
meta-analysis. However, the effect pro-
duced by this bias would be an exag-
geration of the effect. Presence of mis-
classification bias of the exposure to hair
dyes is not unlikely. Because exposure
is generally measured at baseline and
may change during follow-up, the po-
tential for misclassification is higher for
cohort studies, as mentioned by Miller
and Bartsch.
68
If exposure is measured
on a dichotomous scale (exposed vs non-
exposed), the consequence of nondif-
ferential misclassification would be to
bias the estimates toward the absence of
effect. However, globally in our meta-
analysis, we did not detect substantial
differences between pooled results from
case-control and cohort studies.
One possible limitation of our meta-
analysis is that several case-control stud-
ies use the same comparison group
for different outcomes and, therefore,
might not be considered as completely
independent. This raises a multiple com-
parison issue that would result in find-
ing more statistically significant asso-
ciations than they actually exist.
42
This
is particularly true for hematopoietic
cancers because in several studies, the
same control population was used to cal-
culate relative risks for Hodgkin and
non-Hodgkin lymphoma, leukemia, and
multiple myeloma (Table 5).
42,43,46,54,58
This may partially explain the positive
results for hematopoietic cancers. How-
ever, it should be noted that pooling re-
sults from studies that share the same
control group should yield similar re-
sults as pooling results from studies that
use independent control groups, pro-
vided that controls are selected inde-
pendent of the exposure. Restricting our
analysis to studies that used different
comparison sets yielded identical re-
sults (Table 6).
Another possible limitation is that in-
dividual studies may have failed to con-
trol for potential known or unknown
confounders or to take into account po-
tential effect modifiers. A study has
shown that bladder cancer among hair
dye users is restricted to the specific
genotype/phenotype of slow acetyla-
tors of N-acetyltransferase, an enzyme
involved in the detoxification of aro-
matic amines.
4
This and other genetic
factors are potential effect modifiers that
were not addressed in the individual
studies. Furthermore, it is remarkable
that fewer than half of the studies ex-
plicitly adjusted for smoking, an estab-
lished risk factor for cancer, although
some authors decided not to adjust for
it after having ruled out confounding
in their data. Failure to adjust for con-
founders may bias the results in either
direction, toward exaggeration or un-
derestimation of the effect of personal
hair dyes on cancer.
The borderline effect observed for
brain tumors and ovarian cancer is
based on the pooling of only 2 studies
and does not permit a meaningful as-
sessment of the risk. However, these re-
sults are based on reasonably sized case-
control studies, although largely prone
to bias (large proportion of proxy re-
spondents,
60
large number of dead
cases,
60
and imprecise exposure assess-
ment
59
). Future studies may shed more
light on the effect of hair dyes on these
types of tumors.
In conclusion, we did not find strong
evidence of a marked increase in the risk
of cancer among personal hair dye us-
ers. Some aspects related to hematopoi-
etic cancer should be investigated fur-
ther. Efforts should be targeted toward
the assessment of the risk of cancer in
occupational settings where exposure to
hair dyes is more prolonged and has a
higher concentration and frequency than
personal exposure.
Author Contributions: Dr Takkouche had full access
to all of the data in the study and takes responsibility
for the integrity of the data and the accuracy of the
data analysis.
Study concept and design: Takkouche, Etminan.
Acquisition of data: Takkouche, Montes-Martinez.
Analysis and interpretation of data: Takkouche,
Etminan, Montes-Martinez.
Drafting of the manuscript: Takkouche, Montes-
Martinez.
Critical revision of the manuscript for important in-
tellectual content: Etminan.
Statistical analysis: Takkouche, Etminan, Montes-
Martinez.
Administrative, technical, or material support: Montes-
Martinez.
Study supervision: Takkouche.
Financial Disclosures: None reported.
Funding/Support: Dr Etminan is supported by a Ca-
nadian Institutes of Health Research postdoctoral fel-
lowship award.
REFERENCES
1. Ames BN, Kammen HO, Yamasaki E. Hair dyes are
mutagenic: identification of a variety of mutagenic
ingredients. Proc Natl Acad SciUSA. 1975;72:2423-
2427.
2. Sontag JM. Carcinogenicity of substituted-
benzendiamines (phenylenediamines) in rats and mice.
J Natl Cancer Inst. 1981;66:591-602.
3. Silverman DT, Morrison AS, Devesa SS. Bladder
cancer. In: Schottenfeld D, Fraumeni JF, eds. Cancer
Epidemiology and Prevention. New York, NY: Ox-
ford University Press; 1996:1156-1179.
4. Gago-Dominguez M, Castelao JE, Yuan JM, Yu MC,
Ross RK. Use of permanent hair dyes and bladder-
cancer risk. Int J Cancer. 2001;91:575-579.
5. Scientific Committee on Cosmetic Products and
Non-food Products. Opinion Concerning Use of Per-
manent Hair Dyes and Bladder Cancer. Adopted on
April 23, 2004. SCCNFP/0797/04. Available at: http:
//www.europa.eu.int/comm/health/ph_risk
/committees/sccp/documents/out272_en.pdf. Ac-
cessed May 3, 2005.
6. Meadows M. Heading off hair-care disasters: use
caution with relaxers and dyes. FDA Consumer.
January-February 2001. Available at: http://www.fda
.gov/fdac/features/2001/101_hair.html. Accessed De-
cember 29, 2004.
7. Occupational exposures of hairdressers and bar-
bers and personal use of hair colourants. IARC Monogr
Eval Carcinog Risks Hum. 1993;57:43-118.
8. Hair dye cancer alert. BBC News. April 17, 2002.
Available at: http://news.bbc.co.uk/1/hi/health
/1934496.stm. Accessed December 29, 2004.
9. La Vecchia C, Tavani A. Epidemiological evidence
on hair dyes and the risk of cancer in humans. Eur J
Cancer Prev. 1995;4:31-43.
10. Correa A, Jackson L, Mohan A, Perry H, Helzl-
souer K. Use of hair dyes, hematopoietic neoplasms,
and lymphomas: a literature review, I: leukemias and
myelodysplastic syndromes. Cancer Invest. 2000;18:
366-380.
PERSONAL USE OF HAIR DYES AND RISK OF CANCER
2524 JAMA, May 25, 2005—Vol 293, No. 20 (Reprinted) ©2005 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ on 02/25/2013

11. Correa A, Jackson L, Mohan A, Perry H, Helzl-
souer K. Use of hair dyes, hematopoietic neoplasms,
and lymphomas: a literature review, II: lymphomas and
multiple myeloma. Cancer Invest. 2000;18:467-479.
12. Stroup DF, Berlin JA, Morton SC, et al. Meta-
analysis of observational studies in epidemiology: a pro-
posal for reporting. JAMA. 2000;283:2008-2012.
13. Nakata S, Sato J, Ohtake N, Imai K, Yamanaka
H. Epidemiological study of risk factors for bladder can-
cer [in Japanese]. Hinyokika Kiyo. 1995;41:969-977.
14. Hartge P, Hoover R, Altman R, et al. Use of hair
dyes and risk of bladder cancer. Cancer Res. 1982;42:
4784-4787.
15. Bhutta AT, Cleves MA, Casey PH, Cradock MM,
Anand KJ. Cognitive and behavioral outcomes of
school-aged children who were born preterm: a
meta-analysis. JAMA. 2002;288:728-737.
16. Bland JM, Altman DG. Statistical methods for as-
sessing agreement between two methods of clinical
measurement. Lancet. 1986;1:307-310.
17. Takkouche B, Cadarso-Suarez C, Spiegelman D.
Evaluation of old and new tests of heterogeneity in
epidemiologic meta-analysis. Am J Epidemiol. 1999;
150:206-215.
18. Sutton AJ, Abrams KR, Jones DR, Sheldon TA, Song
F. Methods for Meta-analysis in Medical Research.
Chichester, England: John Wiley & Sons; 2000:115.
19. Egger M, Davey Smith G, Schneider M, Minder
C. Bias in meta-analysis detected by a simple, graphi-
cal test. BMJ. 1997;315:629-634.
20. Costa-Bouzas J, Takkouche B, Cadarso-Suarez C,
Spiegelman D. HEpiMA: software for the identifica-
tion of heterogeneity in meta-analysis. Comput Meth-
ods Programs Biomed. 2001;64:101-107.
21. Kinlen LJ, Harris R, Garrod A, Rodriguez K. Use
of hair dyes by patients with breast cancer: a case-
control study. BMJ. 1977;2:366-368.
22. Shore RE, Pasternack BS, Thiessen EU, Sadow M,
Forbes R, Albert RE. A case-control study of hair dye
use and breast cancer. J Natl Cancer Inst. 1979;62:277-
283.
23. Nasca PC, Lawrence CE, Greenwald P, Chorost
S, Arbuckle JT, Paulson A. Relationship of hair dye use,
benign breast disease, and breast cancer. J Natl Can-
cer Inst. 1980;64:23-28.
24. Stavraky KM, Clarke EA, Donner A. Case-
control study of hair dye use by patients with breast
cancer and endometrial cancer. J Natl Cancer Inst.
1979;63:941-945.
25. Stavraky KM, Clarke EA, Donner A. A case-
control study of hair-dye use and cancers of various
sites. Br J Cancer. 1981;43:236-239.
26. Wynder EL, Goodman M. Epidemiology of breast
cancer and hair dyes. J Natl Cancer Inst. 1983;71:481-
488.
27. Koenig KL, Pasternack BS, Shore RE, Strax P. Hair
dye use and breast cancer: a case-control study among
screening participants. Am J Epidemiol. 1991;133:985-
995.
28. Nasca PC, Baptiste MS, Field NA, Metzger BB,
DeMartino R. An epidemiologic case-control study of
breast cancer and exposure to hair dyes. Ann
Epidemiol. 1992;2:577-586.
29. Boice JD Jr, Mandel JS, Doody MM. Breast can-
cer among radiologic technologists. JAMA. 1995;274:
394-401.
30. Cook LS, Malone KE, Daling JR, Voigt LF, Weiss
NS. Hair product use and the risk of breast cancer in
young women. Cancer Causes Control. 1999;10:551-
559.
31. Zheng T, Holford TR, Mayne ST, et al. Use of hair
colouring products and breast cancer risk: a case-
control study in Connecticut. Eur J Cancer. 2002;38:
1647-1652.
32. Petro-Nustas W, Norton ME, al-Masarweh I. Risk
factors for breast cancer in Jordanian women. J Nurs
Scholarsh. 2002;34:19-25.
33. Green A, Willett WC, Colditz GA, et al. Use of
permanent hair dyes and risk of breast cancer. J Natl
Cancer Inst. 1987;79:253-257.
34. Altekruse SF, Henley SJ, Thun MJ. Deaths from he-
matopoietic and other cancers in relation to perma-
nent hair dye use in a large prospective study (United
States). Cancer Causes Control. 1999;10:617-625.
35. Jain M, Morgan RW, Elinson L. Hair dyes and blad-
der cancer. Can Med Assoc J. 1977;117:1131-1133.
36. Neutel CI, Nair RC, Last JM. Are hair dyes asso-
ciated with bladder cancer? Can Med Assoc J. 1978;
119:307-308.
37. Howe GR, Burch JD, Miller AB, et al. Tobacco use,
occupation, coffee, various nutrients, and bladder
cancer. J Natl Cancer Inst. 1980;64:701-713.
38. Ohno Y, Aoki K, Obata K, Morrison AS. Case-
control study of urinary bladder cancer in metropolitan
Nagoya. Natl Cancer Inst Monogr. 1985;69:229-234.
39. Nomura A, Kolonel LN, Yoshizawa CN. Smok-
ing, alcohol, occupation, and hair dye use in cancer
of the lower urinary tract. Am J Epidemiol. 1989;130:
1159-1163.
40. Andrew AS, Schned AR, Heaney JA, Karagas MR.
Bladder cancer risk and personal hair dye use. Int J
Cancer. 2004;109:581-586.
41. Henley SJ, Thun MJ. Use of permanent hair dyes
and bladder-cancer risk. Int J Cancer. 2001;94:903-906.
42. Cantor KP, Blair A, Everett G, et al. Hair dye use
and risk of leukemia and lymphoma. Am J Public
Health. 1988;78:570-571.
43. Zahm SH, Weisenburger DD, Babbitt PA, Saal RC,
Vaught JB, Blair A. Use of hair coloring products and
the risk of lymphoma, multiple myeloma, and chronic
lymphocytic leukemia. Am J Public Health. 1992;82:
990-997.
44. Linos A, Kiamouris C, Foukanelis T, Kourousis C,
Hatzis C, Stefanoudaki K. A case-control study of non-
Hodgkin’s lymphoma. J Epidemiol. 1994;139:S46.
45. Holly EA, Lele C, Bracci PM. Hair-color products
and risk for non-Hodgkin’s lymphoma: a population-
based study in the San Francisco Bay Area. Am J Pub-
lic Health. 1998;88:1767-1773.
46. Miligi L, Seniori Costantini A, Crosignani P, et al.
Occupational, environmental, and life-style factors as-
sociated with the risk of hematolymphopoietic ma-
lignancies in women. Am J Ind Med. 1999;36:60-69.
47. Schroeder JC, Olshan AF, Dent RB, et al. A case-
control study of tobacco use and other non-
occupational risk factors for t(14;18) subtypes of non-
Hodgkin’s lymphoma (United States). Cancer Causes
Control. 2002;13:159-168.
48. Zhang Y, Holford TR, Leaderer B, et al. Hair-
coloring product use and risk of non-Hodgkin’s lym-
phoma: a population-based case-control study in
Connecticut. Am J Epidemiol. 2004;159:148-154.
49. Chiu BC, Weisenburger DD, Zahm SH, et al. Ag-
ricultural pesticide use, familial cancer, and risk of non-
Hodgkin lymphoma. Cancer Epidemiol Biomarkers
Prev. 2004;13:525-531.
50. Grodstein F, Hennekens CH, Colditz GA, Hunter
DJ, Stampfer MJ. A prospective study of permanent
hair dye use and hematopoietic cancer. J Natl Cancer
Inst. 1994;86:1466-1470.
51. Thun MJ, Altekruse SF, Namboodiri MM, Calle
EE, Myers DG, Heath CW Jr. Hair dye use and risk of
fatal cancers in US women. J Natl Cancer Inst. 1994;
86:210-215.
52. Brown LM, Everett GD, Burmeister LF, Blair A. Hair
dye use and multiple myeloma in white men. Am J
Public Health. 1992;82:1673-1674.
53. Herrinton LJ, Weiss NS, Koepsell TD, Daling JR,
Taylor JW, Lyon JL. Exposure to hair-coloring prod-
ucts and the risk of multiple myeloma. Am J Public
Health. 1994;84:1142-1144.
54. Mele A, Szklo M, Visani G, et al. Hair dye use and
other risk factors for leukemia and pre-leukemia: a case-
control study. Am J Epidemiol. 1994;139:609-619.
55. Markovic-Denic L, Jankovic S, Marinkovic J, Rado-
vanovic Z. Brick mortar exposure and chronic lym-
phocytic leukemia. Neoplasma. 1995;42:79-81.
56. Bjork J, Albin M, Welinder H, Tinnerberg H, Mau-
ritzson N, Kauppinen T. Are occupational, hobby, or
lifestyle exposures associated with Philadelphia chro-
mosome positive chronic myeloid leukaemia? Occup
Environ Med. 2001;58:722-727.
57. Rauscher GH, Shore D, Sandler DP. Hair dye use
and risk of adult acute leukemia. Am J Epidemiol. 2004;
160:19-25.
58. Tavani A, Negri E, Franceschi S, Talamini R, Ser-
raino D, La Vecchia C. Hair dye use and risk of lym-
phoid neoplasms and soft tissue sarcomas. Int J Cancer.
2005;113:629-631.
59. Ahlbom A, Navier IL, Norell S, Olin R, Spännare
B. Nonoccupational risk indicators for astrocytomas in
adults. Am J Epidemiol. 1986;124:334-337.
60. Burch JD, Craib KJ, Choi BC, Miller AB, Risch HA,
Howe GR. An exploratory case-control study of brain
tumors in adults. J Natl Cancer Inst. 1987;78:601-609.
61. Holman CDJ, Armstrong BK. Hutchinson’s mela-
notic freckle melanoma associated with non-
permanent hair dyes. Br J Cancer. 1983;48:599-601.
62. Holman CD, Armstrong BK. Hutchinson’s mela-
notic freckle melanoma and the use of non-
permanent hair dyes. Br J Cancer. 1985;52:135.
63. Osterlind A, Tucker MA, Stone BJ, Jensen OM.
The Danish case-control study of cutaneous malig-
nant melanoma, IV: no association with nutritional fac-
tors, alcohol, smoking or hair dyes. Int J Cancer. 1988;
42:825-828.
64. Tzonou A, Polychronopoulou A, Hsieh CC, Re-
belakos A, Karakatsani A, Trichopoulos D. Hair dyes,
analgesics, tranquilizers and perineal talc application
as risk factors for ovarian cancer. Int J Cancer. 1993;
55:408-410.
65. Spitz MR, Fueger JJ, Goepfert H, Newell GR. Sali-
vary gland cancer: a case-coontrol investigation of risk
factors. Arch Otolaryngol Head Neck Surg. 1990;116:
1163-1166.
66. Hennekens CH, Rosner B, Belanger C, Speizer FE,
Bain CJ, Peto R. Use of permanent hair dyes and
cancer among registered nurses. Lancet. 1979;1:
1390-1393.
67. Markowitz JA, Szklo M, Sensenbrenner LL, Warm
S. Hair dyes and acute nonlymphocytic leukemia
[abstract]. Am J Epidemiol. 1985;122:523.
68. Miller AB, Bartsch H. Hair dye use and bladder
cancer. Int J Cancer. 2001;94:901-902.
PERSONAL USE OF HAIR DYES AND RISK OF CANCER
©2005 American Medical Association. All rights reserved. (Reprinted) JAMA, May 25, 2005—Vol 293, No. 20 2525
Downloaded From: http://jama.jamanetwork.com/ on 02/25/2013