International Journal of Epidemiology, 2014, 1825–1834
doi: 10.1093/ije/dyu183
Advance Access Publication Date: 2 September 2014
Original article
Environmental Exposures
Agent Orange exposure and risk of death in
Korean Vietnam veterans: Korean
Veterans Health Study
Sang-Wook Yi,1 So-Yeon Ryu,2* Heechoul Ohrr3 and Jae-Seok Hong4
1
Department of Preventive Medicine and Public Health, Kwandong University College of Medicine,
Gangneung, Republic of Korea, 2Department of Preventive Medicine, Chosun University Medical
School, Gwangju, Republic of Korea, 3Department of Preventive Medicine and Public Health, Yonsei
University College of Medicine, Seoul, Republic of Korea and 4Health Insurance Review and
Assessment Service, Seoul, Republic of Korea
*Corresponding author. Department of Preventive Medicine, Chosun University Medical School, 309 Pilmundae-ro,
Dong-gu, Gwangju, 501–759, Republic of Korea. E-mail: [email protected]
Accepted 8 August 2014
Abstract
Background: Agent Orange (AO) was a mixture of phenoxy herbicides, containing
several dioxin impurities including 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Various
military herbicides, including AO, were sprayed by the US military and allied forces for
military purposes during the Vietnam War. This study was performed to identify the associations between the AO exposure and mortality in Korean Vietnam veterans.
Methods: From 1 January 1992 to 31 December 2005, 180 639 Korean Vietnam veterans
were followed up for vital status and cause of death. The AO exposure index was based
on the proximity of the veteran’s unit to AO-sprayed areas, using a geographical information system-based model. The adjusted hazard ratios and 95% confidence intervals were
calculated by Cox’s proportional hazard model.
Results: The mortality from all causes of death was elevated with AO exposure. The
deaths due to all sites of cancers combined and some specific cancers, including cancers
of the stomach, small intestine, liver, larynx, lung, bladder and thyroid gland, as well as
chronic myeloid leukaemia, were positively associated with AO exposure. The deaths
from angina pectoris, chronic obstructive pulmonary disease and liver disease including
liver cirrhosis were also increased with an increasing AO exposure.
Conclusions: Overall, this study suggests that AO/TCDD exposure may account for mortality from various diseases even several decades after exposure. Further research is
needed to better understand the long-term effects of AO/TCDD exposure on human
health.
Key words: Agent Orange, cohort studies, dioxins, herbicides, Korea, mortality, veterans
C The Author 2014; all rights reserved. Published by Oxford University Press on behalf of the International Epidemiological Association
V
1825
1826
International Journal of Epidemiology, 2014, Vol. 43, No. 6
Key Messages
• Vietnam veterans faced TCDD-contaminated military herbicide (Agent Orange) exposure.
• The mortality from cancers of the stomach, small intestine, liver, larynx, lung, bladder and thyroid gland, and chronic
myeloid leukaemia, was positively associated with Agent Orange exposure.
• The deaths from angina pectoris, chronic obstructive pulmonary disease and liver disease including liver cirrhosis
were increased with increasing Agent Orange exposure.
• Agent Orange/TCDD exposure decades earlier may increase deaths from various diseases.
Introduction
From 1961 to 1971, the United States and allied military
forces sprayed around 77 million litres of herbicides over
the former South Vietnam; Agent Orange (AO) was the
best known and the most used herbicide.1 AO was a
mixture of 2,4-dichlorophenoxyacetic acid (2,4-D) and
2,4,5-trichlorophenoxyacetic acid (2,4,5-T), which contained dioxin contaminants including 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD).2 Despite numerous studies on
the environmental and health effects of AO, the effects of
the chemicals have not been fully explained.3–11
From 1964 to 1973, around 320 000 military personnel
from the Republic of Korea (ROK) were sent to
Vietnam.12 Since the mid 1990s, AO-related health issues
have been investigated in Korean Vietnam veterans; the
relationship between AO and veterans’ health, however,
remains inconclusive.12,13 All-cause mortality in Korean
Vietnam veterans was lower than expected from the general population,14 and this was explained by a ‘healthysoldier effect’, in that Vietnam veterans were more fit and
healthier than the general population or even than nonVietnam veterans.3,4,11,14 In populations with a strong
healthy-worker effect, the mortality and morbidity of AO
exposure among Vietnam veterans should be explored by
comparisons between people with high exposure and veterans with low exposure, rather than using the general
population. Several studies in Korean Vietnam veterans assessed the exposure to AO based on subjective self-report
or crude military information such as the Corps Tactical
Zone in which the veterans’ unit had been stationed.13,15
Therefore, some limitations have been noted.12,13
AO exposure has been assessed primarily by the experience of Vietnam service, and the association between
potential AO exposure and mortality has been examined in
Australian, New Zealand and US Vietnam veterans.3–11
Since the association between AO and morbidity in
Vietnam veterans might vary by ethnicity or geographical
area,7 exploring the association in Korean Vietnam veterans would be helpful to elucidate this. We constructed
the AO exposure index using a geographical information
system (GIS)-based model,16,17 which enabled us to
estimate potential AO exposure at an individual level and
to compare mortality within Vietnam veterans by exposure, rather than using a Vietnam experience-based index.
The aim of this cohort study was to evaluate the association between AO exposure and the risk of death from all
causes as well as specific causes of death among Korean
Vietnam veterans.
Methods
Study population
The Korean Veterans Health Study (KVHS) was established to evaluate primarily the association between
Vietnam War experience and AO exposure, and the morbidities and mortalities from various diseases. In the KVHS
study, from November 1999 to April 2000 the authors
identified 187 897 veterans and were assured of their official residential status as of March 2000 and June 2004.
Since the administrative database of resident registration
could not provide complete information on those deceased
before 1992, and deaths before 1992 were unidentifiable
in the death records of the National Statistical Office, the
study cohort was considered to be established as of 1
January 1992. After excluding 7258 individuals who were
deceased, had emigrated to another country or had an unknown residence before 1992, or whose exposure to AO
was not calculated because of lack of necessary information, finally 180 639 veterans were selected to be followed
up for death.
Follow-up and ascertainment of deaths
The deaths of Korean Vietnam veterans and underlying
causes of their deaths were ascertained by the 1992–2005
death records of the National Statistical Office. Until 31
December 2005, when veterans died the reported date of
death was considered the end of follow-up. When veterans
had emigrated to another country or their residential status
was unknown during 1992–2005, the changed date of
the residential status was considered the date of loss to
International Journal of Epidemiology, 2014, Vol. 43, No. 6
follow-up. A complete follow-up was achieved for 177 899
veterans (98.5%). This study was ethically approved by
the Institutional Review Board of Kwandong University.
Classification of causes of death
The causes of death were classified according to the
International Classification of Diseases 10th revision (ICD10), and we categorized all-causes of death, 15 chapter diseases, 23 specific cancers and 36 specific causes of death
other than cancer. Except for congenital malformations
and skin diseases, generally, causes of death with 10 or
more cases were analysed. Subcategories of leukaemia with
five or more cases were segmented by four-character ICD10 code and included.
AO exposure assessment
The AO exposure index was based on the proximity of the
veteran’s military unit to AO-sprayed areas, using GISbased exposure opportunity model E4.16 Researchers identified the coordinates of the ROK military unit’s post location and tactical area of responsibility during the Vietnam
War.12 The veteran’s deployed unit information was obtained from the Ministry of Defence for the division or brigade level. Each unit’s post location was collected by point
coordinates. Each coordinate representing 1 km by 1 km
within the units’ tactical area of responsibility was obtained. The coordinate information was sent to Stellman’s
team in the USA, and they constructed E4 scores based on
the dates and coordinates.16 Unit-level E4 scores by calendar day were constructed for the Vietnam War period. An
individual E4 score was obtained from the unit in which
the veteran served and the period of deployment. After
adding 1 to each E4 score, the common log-transformed
E4 score (Log10E4) was used as the individual’s AO exposure index. The veterans were categorized into two groups,
low (Log10E4 < 4.0) and high exposure (Log10E4 4). The
exposure index and group classification are describedd
elsewhere in more detail.12,17,18
Statistical analysis
A Cox proportional hazard regression analysis, controlling
for age at cohort entry (as of 1 January 1 1992) and military rank, was implemented. The hazard ratio (HR) for
mortality was calculated using AO exposure as a continuous variable (Log10E4) or by comparing the high-exposure
group with the low-exposure group. The military rank was
included in the model since it could be a partial reflection
of veterans’ socioeconomic status and a confounder for
evaluating Korean veterans’ health.12 Since this veterans’
1827
cohort is a cohort of survivors as of 1992, additional analyses were done in all of the participants with follow-up
until 1998, and in survivors as of 1999 with follow-up
until 2005, with or without stratification by age at enrolment, to evaluate whether the association of AO exposure
differed by follow-up period and age. The P-value was
calculated with two-sided tests. All statistical analyses
were performed using SAS version 9.2 (SAS Institute, Cary,
NC, USA).
Results
The total follow-up person-years summed to 2 403 799,
and 17 529 veterans died during 1992–2005. The average
age (SD) for Vietnam veterans was 46.3(3.5) years as of
1 January 1992. The average age in the low- and highexposure groups was 45.3(3.6) years and 47.5(3.1) years,
respectively. The high-exposure group included more combat units and fewer support units, and had fewer officers
than the low-exposure group. All veterans deployed in
Vietnam from 1971 belonged to the low-exposure group
(Table 1).
The risk of all-cause mortality was elevated with AO
exposure (Table 2). As for ICD-10 chapter diseases, the
risks of deaths from neoplasms, circulatory diseases, digestive diseases and external causes were elevated with an
increase of Log10E4, as well as in the high-exposure group
compared to the low-exposure group. The risk of death
from respiratory disease in the high-exposure group was
higher than in the low-exposure group [1.24, 95% confidence interval (CI) ¼ 1.02–1.50] (Table 2).
As for specific cancers, adjusted HRs (aHRs) of stomach cancer, liver cancer, thyroid cancer and chronic myeloid leukaemia increased as the Log10E4 increased, and
these were also higher in the high-exposure group than in
the low-exposure group (P < 0.05) (Table 3). The aHR of
larynx cancer was elevated with an incremental Log10E4,
and the aHRs of lung cancer (1.15, 95% CI ¼ 1.02–1.30)
and bladder cancer (2.04, 95% CI ¼ 1.17–3.55) were
higher in the high-exposure group than in the low-exposure
group (Table 3).
In specific causes of death other than cancers, the risks
of deaths from angina pectoris, chronic obstructive pulmonary disease (COPD), liver disease, alcoholic liver disease, liver cirrhosis and transport accidents increased with
an increasing Log10E4, and these were also higher in highexposure group compared with the low-exposure group
(P < 0.01) (Table 4).
The associations of AO exposure with mortality were
generally stronger during the earlier period [aHR for allcause (1.16) and cancer mortality (1.20)] of follow-up
than during the later period [aHR for all-cause (1.06) and
1828
International Journal of Epidemiology, 2014, Vol. 43, No. 6
Table 1. Age and Vietnam service characteristics by Agent Orange exposure among Vietnam veterans (n¼180 639)
Characteristics
Age as of 1 January 1992 (years)
Deployed unit
Military rank
Year first deployed to Vietnam
Classification
n
<45
45–49
50–54
55
Capital Division
9th Division
Marine 2nd Brigade
ROK Army Headquarters
Construction Support Group
Naval Transport Group
100th Logistic Command
Unknown
Enlisted
Noncommissioned officer
Company officer
Field officer or general
1966
1967–68
1969–70
1971–
71 014
90 631
12 471
6523
65 550
61 209
5608
5124
9762
565
31 939
882
140 480
25 206
12 209
2744
31 427
49 181
55 725
44 306
%
39.3
50.2
6.9
3.6
36.3
33.9
3.1
2.8
5.4
0.3
17.7
0.5
77.8
14.0
6.8
1.5
17.4
27.2
30.8
24.5
Low exposure
High exposure
n
%
n
%
59 710
27 144
4446
3319
29 906
31 031
2745
5124
6677
504
18 152
480
73 816
12 460
6698
1645
10 324
6166
33 823
44 306
63.1
28.7
4.7
3.5
31.6
32.8
2.9
5.4
7.1
0.5
19.2
0.5
78.0
13.2
7.1
1.7
10.9
6.5
35.7
46.8
11 304
63 487
8025
3204
35 644
30 178
2863
0
3085
61
13 787
402
66 664
12 746
5511
1099
21 103
43 015
21 902
0
13.1
73.8
9.3
3.7
41.4
35.1
3.3
0.0
3.6
0.1
16.0
0.5
77.5
14.8
6.4
1.3
24.5
50.0
25.5
0.0
P-value
<0.001
<0.001
<0.001
<0.001
ROK, Republic of Korea.
Table 2. Numbers of deaths and adjusteda hazard ratios for major causes of death according to Agent Orange exposure
Causes of death
All causes of death
Infectious and parasitic diseases
Neoplasms
Diseases of the blood and
blood-forming organs
Endocrine diseases
Mental disorders
Diseases of the nervous system
Diseases of the circulatory system
Diseases of the respiratory system
Diseases of the digestive system
Diseases of the skin
Diseases of the musculoskeletal system
Diseases of the genitourinary system
Congenital malformations
Symptoms, not elsewhere classified
External causes of morbidity
ICD-10
A00–R99,
V01–Y89
A00–B99
C00–D48
D50–D89
E00–E88
F01–F99
G00–G98
I00–I99
J00–J98
K00–K92
L00–L98
M00–M99
N00–N98
Q00–Q99
R00–R99
V01–Y89
Log10E4b
Total
deaths
n
Low
Exposure
Deaths
n
HRc
95% CI
P-value
17 529
1.03
1.02–1.04
377
6282
24
1.00
1.03
1.13
746
237
140
3180
446
2212
9
38
190
1
450
3197
1.02
1.03
0.98
1.02
1.04
1.05
1.03
0.89
0.98
0.78
1.04
1.04
High exposure
Deaths
n
HRd
95% CI
P-value
<0.001
7973
9556
1.10
1.07–1.14
<0.001
0.95–1.05
1.02–1.04
0.91–1.39
0.894
<0.001
0.266
181
2803
10
196
3479
14
1.04
1.13
1.29
0.84–1.29
1.07–1.19
0.56–2.95
0.690
<0.001
0.548
0.98–1.06
0.96–1.10
0.91–1.07
1.00–1.04
0.99–1.09
1.03–1.08
0.74–1.44
0.76–1.03
0.92–1.05
0.29–2.09
1.00–1.09
1.02–1.06
0.287
0.434
0.709
0.028
0.084
<0.001
0.842
0.108
0.590
0.617
0.077
<0.001
347
113
68
1464
180
1005
4
22
92
1
210
1473
399
124
72
1716
266
1207
5
16
98
0
240
1724
1.00
1.07
0.95
1.04
1.24
1.18
1.21
0.57
0.95
0.00
1.09
1.18
0.86–1.16
0.82–1.40
0.68–1.34
0.97–1.12
1.02–1.50
1.08–1.29
0.31–4.78
0.30–1.09
0.71–1.27
0.987
0.620
0.766
0.289
0.032
<0.001
0.781
0.091
0.719
0.90–1.33
1.09–1.27
0.381
<0.001
CI, confidence interval; HR, hazard ratio; ICD-10, 10th revision of the International Classification of Diseases.
a
Adjusted for age at cohort entry (as of 1 January 1992) and military rank during service in Vietnam.
b
Log10E4 is the common log-transformed E4 score which was constructed using a geographical information system-based model grounded in the proximity of
the veteran’s military unit to an Agent Orange-sprayed area.
c
Hazard ratio as 1-point increase in Log10E4.
d
Compared with the low-exposure group (2-group analysis).
International Journal of Epidemiology, 2014, Vol. 43, No. 6
1829
Table 3. Numbers of deaths and adjusteda hazard ratios for death from malignant neoplasms according to Agent Orange
exposure
Causes of death
Neoplasms
Oral cavity cancer
Oesophagus cancer
Stomach cancer
Small intestine cancer
Colorectal cancer
Liver cancer
Gall bladder cancer
Pancreas cancer
Larynx cancer
Lung cancer
Thymus cancer
Bone cancer
Malignant melanoma
Prostate cancer
Renal cancer
Bladder cancer
Central nervous system cancer
Thyroid cancer
Non-Hodgkin lymphoma
Multiple myeloma
Leukaemia
Acute lymphoblastic leukaemia
Acute myeloblastic leukaemia
Chronic myeloid leukaemia
ICD10
C00–D48
C00–C14
C15
C16
C17
C18–C21
C22
C23–C24
C25
C32
C33–C34
C37
C40–C41
C43
C61
C64–C66
C67
C70–C72
C73
C82–C85
C90
C91–C95
C91.0
C92.0
C92.1
Log10E4b
Total
deaths
n
HRc
95% CI
P-value
82
162
1077
19
366
2053
215
255
82
1170
10
16
11
38
63
61
73
11
103
39
107
5
46
15
1.05
1.02
1.05
1.11
1.02
1.03
1.05
1.03
1.13
1.02
0.99
0.81
1.26
0.94
0.99
1.13
1.00
2.88
1.04
0.98
1.04
0.86
1.02
1.55
0.94–1.17
0.94–1.10
1.02–1.08
0.87–1.40
0.97–1.07
1.00–1.05
0.98–1.12
0.97–1.09
1.00–1.28
0.99–1.05
0.72–1.36
0.64–1.04
0.89–1.77
0.81–1.09
0.88–1.12
0.98–1.29
0.90–1.12
1.12–7.39
0.95–1.15
0.85–1.14
0.95–1.14
0.57–1.28
0.89–1.18
1.06–2.27
0.405
0.590
0.001
0.403
0.458
0.023
0.159
0.367
0.044
0.185
0.947
0.097
0.187
0.418
0.931
0.084
0.986
0.028
0.401
0.833
0.403
0.453
0.748
0.024
Low
exposure
Deaths
n
High exposure
Deaths
n
HRd
95% CI
P-value
37
64
464
5
179
946
95
114
32
497
6
11
5
21
33
19
37
1
47
20
49
3
22
2
45
98
613
14
187
1,107
120
141
50
673
4
5
6
17
30
42
36
10
56
19
58
2
24
13
1.07
1.26
1.17
2.88
0.96
1.12
1.19
1.15
1.28
1.15
0.74
0.48
1.49
0.68
0.88
2.04
0.88
11.31
1.18
0.81
1.18
0.66
1.17
7.91
0.68–1.68
0.91–1.75
1.03–1.33
1.00–8.28
0.78–1.19
1.02–1.23
0.90–1.57
0.89–1.48
0.80–2.03
1.02–1.30
0.19–2.92
0.16–1.49
0.41–5.40
0.36–1.31
0.53–1.47
1.17–3.55
0.55–1.40
1.33–96.55
0.79–1.77
0.43–1.54
0.80–1.76
0.11–4.10
0.64–2.15
1.67–37.52
0.773
0.159
0.014
0.050
0.723
0.013
0.225
0.291
0.298
0.019
0.669
0.206
0.543
0.248
0.629
0.012
0.584
0.027
0.415
0.527
0.403
0.658
0.609
0.009
CI, confidence interval; HR, hazard ratio; ICD-10, 10th revision of the International Classification of Diseases.
a
Adjusted for age at cohort entry (as of 1 January 1992) and military rank during service in Vietnam.
b
Log10E4 is the common log-transformed E4 score which was constructed using a geographical information system-based model grounded in the proximity of
the veteran’s military unit to an Agent Orange-sprayed area.
c
Hazard ratio as 1-point increase in Log10E4.
d
Compared with the low-exposure group (2-group analysis).
cancer mortality (1.09)] when comparing high-exposure
group with low-exposure group, and the associations also
were generally stronger at younger age (below 45 years)
than they were at older age (Supplementary Tables 1–3,
available as Supplementary data at IJE online).
Discussion
All causes of death and all cancers combined
The results showed that a higher AO exposure modestly
elevated the risk of all-cause death. Some studies, in which
Vietnam veterans were compared with general population,
have clearly shown the well-known ‘healthy-soldier effect’,
that Vietnam veterans had a lower all-cause mortality than
general population.7,11,14 The present study, by comparing
the mortality of veterans with that of comparable controls
according to AO exposure, may have provided evidence of
the toxic effects of AO in accord with studies in US
veterans.4,6,10
TCDD has been classified as a carcinogen to humans
for all sites of cancer by the International Agency for
Research on Cancer (IARC).19 This study showed that the
mortality from all cancers combined was increased with
TCDD-contaminated AO exposure, which was in line with
previous studies regarding occupationally TCDD-exposed
workers.20–23
Gastrointestinal cancers
Gastrointestinal cancers rarely have been associated with
AO/TCDD exposure in previous cohort studies,24 but
1830
International Journal of Epidemiology, 2014, Vol. 43, No. 6
Table 4. Numbers of deaths and adjusteda hazard ratios for specific causes of death according to Agent Orange exposure
Causes of death
Infectious and parasitic diseases
Tuberculosis
Septicaemia
Viral hepatitis
Endocrine diseases
Diabetes mellitus
Mental disorders
Mental disorders due to psychoactive
substance use
Schizophrenia
Diseases of the nervous system
Spinal muscular atrophy
Parkinson disease
Alzheimer disease
Epilepsy
Diseases of the circulatory system
Hypertension
Ischaemic heart diseases
Angina pectoris
Acute myocardial infarction
Chronic ischaemic heart disease
Other heart disease
Cerebrovascular diseases
Subarachnoid haemorrhage
Intracerebral haemorrhage
Cerebral infarction
Stroke, not specified
Aortic aneurysm
Diseases of the respiratory system
Pneumonia
Chronic lower respiratory diseases
Chronic obstructive pulmonary diseases
Asthma
Diseases of the digestive system
Peptic ulcer
Liver diseases
Alcoholic liver disease
Liver cirrhosis
Diseases of the genitourinary system
Acute renal failure
Chronic renal failure
Symptoms, not elsewhere classified
Other sudden death, cause unknown
External causes of morbidity
Transport accidents
Falls
Accidental drowning
Exposure to fire
Accidental poisoning
Intentional self-harm
Assault
ICD10
A00–B99
A15–A19
A40–A41
B15–B19
E00–E88
E10–E14
F01–F99
F10–F19
F20
G00–G98
G12
G20–G21
G30
G40–G41
I00–I99
I10–I13
I20–I25
I20
I21
I25
I26–I51
I60–I69
I60
I61–I62
I63
I64
I71
J00–J98
J12–J18
J40–J47
J40–J44, J47
J45–J46
K00–K92
K25–K27
K70–K76
K70
K74
N00–N98
N17
N18
R00–R99
R96
V01–Y89
V01–V99
W00–W19
W65–W74
X00–X09
X40–X49
X60–X84
X85–Y09
Log10E4b
Total
deaths
n
Low
exposure
Deaths
n
HRc
95% CI
Deaths
n
HRd
95% CI
227
71
58
1.00
1.01
0.99
0.94–1.07
0.90–1.13
0.87–1.13
0.974
0.877
0.860
105
36
31
122
35
27
1.12
0.89
0.93
0.85–1.47
0.55–1.45
0.53–1.64
0.424
0.643
0.810
703
1.02
0.98–1.05
0.414
327
376
0.99
0.85–1.15
0.856
203
1.03
0.96–1.11
0.388
97
106
1.09
0.81–1.46
0.574
9
0.97
0.69–1.35
0.838
6
3
0.46
0.10–2.01
0.301
17
25
6
23
0.94
1.01
0.87
1.14
0.74–1.18
0.83–1.22
0.60–1.25
0.91–1.41
0.591
0.947
0.459
0.255
9
12
3
10
8
13
3
13
0.80
0.88
0.86
1.70
0.29–2.15
0.40–1.95
0.17–4.31
0.69–4.20
0.651
0.752
0.857
0.251
192
843
60
699
82
439
1618
117
758
309
272
29
1.06
0.99
1.22
0.98
0.97
1.07
1.01
1.02
1.03
0.99
1.01
1.05
0.99–1.14
0.96–1.03
1.05–1.41
0.95–1.02
0.88–1.08
1.02–1.12
0.99–1.04
0.93–1.12
0.99–1.07
0.94–1.05
0.95–1.07
0.88–1.26
0.108
0.729
0.007
0.313
0.585
0.007
0.353
0.677
0.122
0.848
0.753
0.572
82
406
17
347
41
207
739
57
351
142
121
12
110
437
43
352
41
232
879
60
407
167
151
17
1.18
0.99
2.34
0.93
0.92
1.10
1.01
1.08
1.05
0.96
1.00
1.26
0.88–1.58
0.86–1.14
1.32–4.15
0.80–1.09
0.59–1.44
0.91–1.35
0.92–1.12
0.74–1.58
0.90–1.22
0.76–1.20
0.78–1.27
0.59–2.66
0.278
0.897
0.003
0.383
0.727
0.322
0.785
0.691
0.538
0.701
0.986
0.554
107
190
115
75
1.00
1.07
1.13
1.00
0.91–1.10
0.99–1.15
1.02–1.26
0.89–1.12
0.979
0.085
0.016
0.991
50
74
35
39
57
116
80
36
1.02
1.29
1.73
0.88
0.69–1.50
0.95–1.73
1.16–2.60
0.55–1.42
0.940
0.098
0.008
0.608
38
2009
406
1453
0.98
1.05
1.11
1.05
0.84–1.15
1.03–1.08
1.05–1.17
1.02–1.08
0.826
<0.001
<0.001
<0.001
19
913
170
666
19
1096
236
787
0.92
1.19
1.43
1.17
0.47–1.79
1.09–1.31
1.16–1.77
1.05–1.30
0.800
<0.001
<0.001
0.005
16
134
0.98
0.98
0.77–1.25
0.90–1.06
0.887
0.546
7
69
9
65
1.05
0.87
0.38–2.92
0.61–1.24
0.920
0.439
49
1.07
0.92–1.24
0.394
23
26
1.36
0.73–2.52
0.328
1102
293
101
62
85
783
78
1.06
1.06
0.98
1.03
1.11
1.01
1.07
1.03–1.09
1.00–1.13
0.89–1.08
0.91–1.18
0.99–1.25
0.98–1.05
0.95–1.20
<0.001
0.050
0.623
0.614
0.071
0.421
0.283
511
120
57
25
34
367
37
591
173
44
37
51
416
41
1.21
1.37
0.83
1.48
1.45
1.11
1.27
1.07–1.37
1.07–1.76
0.55–1.26
0.86–2.55
0.92–2.30
0.96–1.29
0.78–2.06
0.003
0.014
0.387
0.159
0.111
0.168
0.328
P-value
High exposure
P-value
CI, confidence interval; HR, hazard ratio; ICD-10, 10th revision of the International Classification of Diseases.
a
Adjusted for age at cohort entry (as of 1 January 1992) and military rank during service in Vietnam.
b
Log10E4 is the common log-transformed E4 score which was constructed using a geographical information system-based model grounded in the proximity of
the veteran’s military unit to an Agent Orange-sprayed area.
c
Hazard ratio as 1-point increase in Log10E4.
d
Compared with the low-exposure group (2-group analysis).
International Journal of Epidemiology, 2014, Vol. 43, No. 6
several case-control studies have reported a positive association mainly with stomach cancer.25 The association of
stomach cancer in Western populations has been controversial. However, the carcinogenic effects of AO/TCDD exposure on the stomach may be more apparent in the Korean
population where stomach cancer is the most common cancer. As shown in the present study, previous research among
Korean rural populations also found that high exposure to
herbicides (and pesticides) increased cancer incidence of the
digestive organs (aHR ¼ 1.5, 95% CI ¼ 1.0–2.3) including
stomach (aHR ¼ 1.6, 95% CI ¼ 0.9–2.8) and liver
(aHR ¼ 2.0, 95% CI ¼ 0.7–5.9) cancer.26 However, the possibility that the low excess risk of stomach cancer in the present study might be accounted for uncontrolled confounding
by important risk factors such as Helicobacter pylori,
smoking and diet should not be overlooked. Meanwhile, the
present study showed that the death rate from small intestine
cancer was higher in the high-exposure group (aHR ¼ 2.88,
95% CI ¼ 1.00–8.28); the association of small intestine cancer has rarely been examined in previous studies.20,27
Liver cancer and liver diseases
Few studies have reported the positive association between
AO/TCDD exposure and hepatobiliary cancers in Vietnam
veterans or among occupational cohorts,28,29 whereas
TCDD is an established carcinogen in animal models.30
Inconsistent hepatotoxicity in previous human studies has
been explained by various reasons such as the lack of good
measures of exposure, a small number of subjects or not
applying dose-response types of analysis.31 However, the
present study revealed a meaningful association between
AO exposure and deaths from liver cancer, in line with
another Korean study,26 based in part on a large number
of deaths from liver cancer32 and a dose-response analysis
of individual exposure assessment.
The hepatotoxicity of TCDD is manifested by hypertrophy, necrosis, fatty degeneration, inflammation, hepatic
lipid peroxidation and porphyria of liver cells.31,33–36 The
prevalence of liver diseases including liver cirrhosis has
been associated with possible AO exposure in Vietnam veterans,2,37 whereas mortality from liver diseases has rarely
been associated with AO.4,6,24 The current study showed
that the risk of deaths from liver diseases including alcoholic liver diseases and liver cirrhosis was elevated with
AO exposure. The prevalences of chronic hepatic viral infection including hepatitis B and C and alcohol drinking,
known risk factors of hepatocellular carcinoma and liver
cirrhosis, are much higher in Korea than in Western countries.38 The hepatotoxicity of AO/TCDD might be confounded by chronic hepatic infection and alcohol drinking,
which were not adjusted for in the present study. Further
1831
research may reveal the mechanism of hepatotoxic effects
of AO in more detail.
Other sites of cancer
Our study found a modestly higher mortality from lung
cancer with AO exposure, in accordance with previous
research,19 whereas our results did not find association between potential TCDD exposure and soft tissue sarcoma
(ICD-10, C47 þ C49: aHR ¼ 0.96, 95% CI ¼ 0.33–2.80 in
high-exposure group compared to low-exposure group)
that was suggested in previous studies.19,24
The risk of death from bladder cancer in the highexposure group was higher than in the low-exposure group
in this study. Even though bladder cancer is the most
common genitourinary cancer, the relationship between
occupational and environmental exposure to dioxinrelated chemicals and bladder cancer has not been fully
elucidated.29,39–41 Meanwhile cacodylic acid, which was
one of the organo-arsenic compounds in military herbicides, has been known to induce bladder tumours in animal experiments,42 and exposure to arsenic has been
linked with bladder cancer.43
This study found that a greater exposure to AO was significantly related to a greater mortality from thyroid cancer. The more the TCDD contamination in the residential
area, the higher the incidence of thyroid cancer was in
Seveso residents (P > 0.05).41 In animal experiments, 2,4D and 2,4,5-T were suggested to be mutagenic or carcinogenic to the thyroid,44–46 but conclusive evidence of AO/
TCDD as a carcinogen to the human thyroid has been lacking. Further research into the histological type of thyroid
cancer may elucidate the effects of AO-related chemicals
on the thyroid gland.
Non-Hodgkin lymphoma (NHL) has been tentatively
associated with TCDD exposure.19 In the present study,
the mortality from NHL was modestly higher in the highexposure group (aHR ¼ 1.2, 95% CI ¼ 0.8–1.8) than in the
low-exposure group, whereas the mortality from chronic
myeloid leukaemia was strongly elevated with AO exposure (aHR ¼ 7.9, 95% CI ¼ 1.7–37.5 in the high-exposure
group). Few prospective studies have examined the association between AO/TCDD exposure and leukaemia and
the results have been inconclusive.24 Our results may
indicate new grounds for suggesting carcinogenic effects of
AO/TCDD exposure on myeloid leukaemia. Further research is needed to confirm this association.
Non-cancerous causes of death other
than liver diseases
In this study, the mortality from angina and other heart
diseases was associated with AO exposure but deaths from
1832
hypertension and myocardial infarction were not associated with AO exposure. A few studies have reported significant associations between AO-related chemicals and
circulatory diseases, including hypertension and ischaemic
heart disease, but the association remains unclear.4
Therefore further investigation is needed.
The risk of death from chronic obstructive pulmonary
disease (COPD) had a positive relationship with AO exposure (aHR ¼ 1.73, 95% CI ¼ 1.16–2.60 in the highexposure group compared with the low-exposure group).
The association between environmental and occupational
exposure to AO/TCDD and non-malignant respiratory diseases has been reported in previous research,4,24,45 but the
findings have not been consistent in Vietnam veterans.2,3,6,11
However, the present study did not adjust for smoking, an
important risk factor for COPD. Our results, nevertheless,
could provide valuable evidence of the harmful effects of
AO/TCDD to lung tissue. Mortality from asthma, however,
was not associated with AO exposure. The different inflammatory processes affecting COPD and asthma might result
in different associations with AO exposure.46
An incremental exposure to AO led to a increased risk of
death from external causes, especially transport accidents
and falls. Vietnam veterans, compared with non-Vietnam
veterans or general population, had higher mortality from
some external causes such as transport accidents, intoxication and suicide.3,10 This can be explained by the illegal use
of drugs and posttraumatic stress disorder (PTSD), both of
which were related to the war experience. However, current
research could not confirm whether high-exposure group
had more severe war experience, PTSD or drug addiction
than the low-exposure group.
Limitations and strengths of the study
This study, however, has some limitations. First, our cohort was constructed among survivors as of 1 January
1992, when at least 19 years had passed since the veterans
had returned from Vietnam. Our results showed that HRs
of AO exposure were generally higher during the earlier
period of follow-up than during the later period, and they
also were generally higher in veterans aged below 45 years
than in those aged 45 and above at enrolment
(Supplementary Tables 1–3, available as Supplementary
data at IJE online). Furthermore, among younger veterans
less than 45 years old, aHRs for leukaemia and NHL, two
cancers with short latency, were higher in [2.4 (95%
CI ¼ 1.1–5.4) and 1.8 (95% CI ¼ 0.8–4.4), respectively]
the high-exposure group relative to the low-exposure
group; whereas among veterans aged 45 years or above as
of 1992, they were not higher [1.04 (95% CI ¼ 0.6–1.7)
and 0.96 (95% CI ¼ 0.6–1.6), respectively]. Therefore,
International Journal of Epidemiology, 2014, Vol. 43, No. 6
there is a possibility that vulnerable veterans with high AO
exposure and subsequent consequences may have died before 1992, at a young age. In such cases, the risk of AO exposure may be underestimated in this study,14 especially
for diseases with short latency and high fatality such as
lymphohaematopoietic malignancies.
Second, the present assessment of AO exposure has some
limitations in accuracy and precision, especially compared
with other studies in which exposure was validated with levels of target chemicals in human tissue.12,17,18 However, we
believe that this GIS-based exposure index is more valid and
reliable than assessments based on subjective self-report or
Vietnam experience.12,18 On the other hand, non-differential misclassification of the exposure index was possible in
this exposure assessment, which may have biased the estimation of the hazard ratios toward the null.
Third, some important risk factors of death, such as
smoking (e.g. for lung cancer and COPD), drinking (e.g.
for liver cancer and liver diseases) and obesity (e.g. for cardiovascular diseases and diabetes) were not adjusted for.
However neither smoking, alcohol drinking nor obesity
were prevalent in the high-exposure group compared with
the low-exposure group in Korean Vietnam veterans,12
and we believe that not adjusting for those variables in this
study would not overestimate the effects of the AO exposure. Nonetheless, not adjusting for those variables as well
as other specific risk factors for some causes of death (e.g.
H. pylori infection for stomach cancer, and infection with
chronic hepatitis virus and liver fluke for liver cancer) is a
potential limitation of the study.
Fourth, information on occupational or environmental
exposure to herbicides/TCDD, such as use of herbicides
and food consumption, was not controlled for.
Fifth, as a mortality study, our results may be different
from those of incidence or prevalence studies, especially in
diseases with low severity and long survival time, since the
severity of diseases and treatment-related factors may affect the estimated risk of death from some diseases.41
Despite these limitations, this study has several
strengths, including a large-sized cohort, a nearly complete
follow-up over a long period and the use of an internal
comparison group by exposure level rather than with the
general population. Therefore, our results may provide
meaningful evidence for the effects of AO/TCDD exposure
on human health, such as deaths from various cancers, angina, COPD and liver diseases, some of which have rarely
been examined.
Conclusion
This study showed that the mortality from all cancers combined was positively associated with AO exposure, which
International Journal of Epidemiology, 2014, Vol. 43, No. 6
generally supports TCDD as a human carcinogen. The
mortality from cancers of the stomach, liver, larynx, lung,
bladder and thyroid gland, and chronic myeloid leukaemia, was also positively associated with AO exposure. An
elevated mortality from COPD with an increasing AO exposure was found and generally concurs with the potential
effect of TCDD on lung tissue. A subtle excess mortality
from liver cirrhosis, as well as the aforementioned liver
cancer with AO exposure, was observed. However, the
clear interpretation of some results may be limited for the
various reasons that readers should note. Further research
is needed to better understand the long-term effects of
AO/TCDD exposure on human health.
Supplementary Data
Supplementary data are available at IJE online.
Funding
This work was supported by a research grant from the Korean
Ministry of Patriots and Veterans Affairs (1999, 2003 and 2009).
Acknowledgements
The authors greatly thank J.M. Stellman and S.D. Stellman for their
permission to use their herbicide exposure data. The authors truly
thank the staff of the Korean National Statistical Office for the mortality data.
Specific author contributions
Drafting the manuscript (including checking the references) and
guarantor for the paper: S.Y and S.R.; performing statistical analysis
and interpretation of the results: S.Y.; reviewing the literature and
revising the manuscript critically: H.O.; reviewing the literatures, interpreting the results and editing the manuscripts: J.H.; all the authors approved the final draft.
Conflict of interest: None declared.
References
1. Stellman JM, Stellman SD, Christian R, Weber T, Tomasallo C.
The extent and patterns of usage of Agent Orange and other
herbicides in Vietnam. Nature 2003;422:681–87.
2. Kang HK, Dalager NA, Needham LL et al. Health status of
Army Chemical Corps Vietnam veterans who sprayed defoliant
in Vietnam. Am J Ind Med 2006;49:875–84.
3. Boehmer TK, Flanders WD, McGeehin MA, Boyle C, Barrett
DH. Postservice mortality in Vietnam veterans: 30-year followup. Arch Intern Med 2004;164:1908–16.
4. Cypel Y, Kang H. Mortality patterns of Army Chemical Corps
veterans who were occupationally exposed to herbicides in
Vietnam. Ann Epidemiol 2010;20:339–46.
5. Dalager NA, Kang HK. Mortality among Army Chemical Corps
Vietnam veterans. Am J Ind Med 1997;31:719–26.
1833
6. Ketchum NS, Michalek JE. Postservice mortality of Air Force veterans occupationally exposed to herbicides during the Vietnam
War: 20-year follow-up results. Mil Med 2005;170:406–13.
7. McBride D, Cox B, Broughton J, Tong D. The mortality and cancer experience of New Zealand Vietnam war veterans: a cohort
study. BMJ Open 2013;3:e003379.
8. Michalek JE, Ketchum NS, Akhtar FZ. Postservice mortality of
US Air Force veterans occupationally exposed to herbicides in
Vietnam: 15-year follow-up. Am J Epidemiol 1998;148:786–92.
9. Visintainer PF, Barone M, McGee H, Peterson EL.
Proportionate mortality study of Vietnam-era veterans of
Michigan. J Occup Environ Med 1995;37:423–28.
10. Watanabe KK, Kang HK. Military service in Vietnam and the
risk of death from trauma and selected cancers. Ann Epidemiol
1995;5:407–12.
11. Wilson EJ, Horsley KW, van der Hoek R. Australian Vietnam
Veterans Mortality Study 2005. Canberra: Department of
Veterans’ Affairs, 2005.
12. Yi SW, Ohrr H, Hong JS, Yi JJ. Agent Orange exposure and
prevalence of self-reported diseases in Korean Vietnam veterans.
J Prev Med Public Health 2013;46:213–25.
13. Kim JS, Lim HS, Cho SI, Cheong HK, Lim MK. Impact of Agent
Orange exposure among Korean Vietnam veterans. Ind Health
2003;41:149–57.
14. Yi SW, Ohrr HC. Mortality patterns among Korean Vietnam
veterans. Korean J Occup Environ Med 2011;23:279–86.
15. Kim HA, Kim EM, Park YC et al. Immunotoxicological effects
of Agent Orange exposure to the Vietnam War Korean veterans.
Ind Health 2003;41:158–66.
16. Stellman JM, Stellman SD, Weber T, Tomasallo C, Stellman AB,
Christian R Jr. A geographic information system for characterizing exposure to Agent Orange and other herbicides in Vietnam.
Environ Health Perspect 2003;111:321–28.
17. Yi SW, Ohrr H, Won JU, Song JS, Hong JS. Serum 2,3,7,8-tetrachlorodibenzo-p-dioxin levels and their association with age,
body mass index, smoking, military record-based variables, and
estimated exposure to Agent Orange in Korean Vietnam veterans. J Prev Med Public Health 2013;46:226–36.
18. Yi SW, Hong JS, Ohrr H, Yi JJ. Agent Orange exposure and disease prevalence in Korean Vietnam veterans: The Korean
Veterans Health Study. Environ Res 2014;133C:56–65.
19. Baan R, Grosse Y, Straif K et al. A review of human carcinogens.
Part F: chemical agents and related occupations. Lancet Oncol
2009;10:1143–44.
20. Fingerhut MA, Halperin WE, Marlow DA et al. Cancer mortality in workers exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin.
N Engl J Med 1991;324:212–18.
21. Kogevinas M, Becher H, Benn T et al. Cancer mortality in workers exposed to phenoxy herbicides, chlorophenols, and dioxins.
An expanded and updated international cohort study. Am J
Epidemiol 1997;145:1061–75.
22. Hooiveld M, Heederik DJ, Kogevinas M et al. Second follow-up
of a Dutch cohort occupationally exposed to phenoxy herbicides, chlorophenols, and contaminants. Am J Epidemiol
1998;147:891–901.
23. Ott MG, Zober A. Cause specific mortality and cancer incidence
among employees exposed to 2,3,7,8-TCDD after a 1953 reactor
accident. Occup Environ Med 1996;53:606–12.
1834
24. Consonni D, Pesatori AC, Zocchetti C et al. Mortality in a population exposed to dioxin after the Seveso, Italy, accident in 1976:
25 years of follow-up. Am J Epidemiol 2008;167:847–58.
25. Mills PK, Yang RC. Agricultural exposures and gastric cancer
risk in Hispanic farm workers in California. Environ Res
2007;104:282–89.
26. Sull JW, Yi SW, Sohn TY, Jee SH, Nam CM, Ohrr H. Pesticides
and cancer incidence. Korean J Prev Med 2002;35:24–32.
27. Steenland K, Piacitelli L, Deddens J, Fingerhut M, Chang LI.
Cancer, heart disease, and diabetes in workers exposed to
2,3,7,8-tetrachlorodibenzo-p-dioxin. J Natl Cancer Inst
1999;91:779–86.
28. Collins JJ, Strauss ME, Levinskas GJ, Conner PR. The mortality
experience of workers exposed to 2,3,7,8-tetrachlorodibenzo-pdioxin in a trichlorophenol process accident. Epidemiology
1993;4:7–13.
29. McBride DI, Collins JJ, Humphry NF et al. Mortality in workers
exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin at a trichlorophenol plant in New Zealand. J Occup Environ Med
2009;51:1049–56.
30. Walker NJ, Wyde ME, Fischer LJ, Nyska A, Bucher JR.
Comparison of chronic toxicity and carcinogenicity of 2,3,7,8tetrachlorodibenzo-p-dioxin (TCDD) in 2-year bioassays in female Sprague-Dawley rats. Mol Nutr Food Res
2006;50:934–44.
31. Michalek JE, Ketchum NS, Longnecker MP. Serum dioxin and
hepatic abnormalities in veterans of Operation Ranch Hand.
Ann Epidemiol 2001;11:304–11.
32. Jung KW, Won YJ, Park S et al. Cancer statistics in Korea: incidence, mortality and survival in 2005. J Korean Med Sci
2009;24:995–1003.
33. Al-Bayati ZA, Stohs SJ. The role of iron in 2,3,7,8-tetrachlorodibenzo-p-dioxin-induced lipid peroxidation by rat liver microsomes. Toxicol Lett 1987;38:115–21.
34. Brooks AL, Jordan SW, Bose KK, Smith J, Allison DC. The cytogenetic and hepatotoxic effects of dioxin on mouse liver cells.
Cell Biol Toxicol 1988;4:31–40.
35. Goldstein JA, Hickman P, Bergman H, Vos JG. Hepatic porphyria induced by 2,3,7,8-tetrachlorodibenzo-p-dioxin in the
mouse. Res Commun Chem Pathol Pharmacol 1973;6:919–28.
International Journal of Epidemiology, 2014, Vol. 43, No. 6
36. Shen ES, Gutman SI, Olson JR. Comparison of 2,3,7,8tetrachlorodibenzo-p-dioxin-mediated
hepatotoxicity
in
C57BL/6J and DBA/2J mice. J Toxicol Environ Health
1991;32:367–81.
37. O’Toole BI, Marshall RP, Grayson DA et al. The Australian
Vietnam Veterans Health Study: II. self-reported health of veterans compared with the Australian population. Int J Epidemiol
1996;25:319–30.
38. Park KW, Park JW, Choi JI et al. Survival analysis of 904 patients with hepatocellular carcinoma in a hepatitis B virus-endemic area. J Gastroenterol Hepatol 2008;23:467–73.
39. Boers D, Portengen L, Bueno-de-Mesquita HB, Heederik D,
Vermeulen R. Cause-specific mortality of Dutch chlorophenoxy
herbicide manufacturing workers. Occup Environ Med
2010;67:24–31.
40. Collins JJ, Bodner K, Aylward LL, Wilken M, Bodnar CM.
Mortality rates among trichlorophenol workers with exposure
to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Am J Epidemiol
2009;170:501–06.
41. Pesatori AC, Consonni D, Rubagotti M, Grillo P, Bertazzi PA.
Cancer incidence in the population exposed to dioxin after the
‘Seveso accident’: twenty years of follow-up. Environ Health
2009;8:39.
42. Cohen SM, Arnold LL, Eldan M, Lewis AS, Beck BD.
Methylated arsenicals: the implications of metabolism and carcinogenicity studies in rodents to human risk assessment. Crit
Rev Toxicol 2006;36:99–133.
43. Straif K, Benbrahim-Tallaa L, Baan R et al. A review of human
carcinogens. Part C: metals, arsenic, dusts, and fibres. Lancet
Oncol 2009;10:453–54.
44. National Toxicology Program. Carcinogenesis Bioassay of
2,3,7,8-Tetrachlorodibenzo-p-dioxin (CAS No. 1746–01–6) in
Osborne-Mendel Rats and B6c3F1 Mice (Gavage Study).
Bethesda, MD: National Toxicology Program, 1982.
45. Bertazzi PA, Bernucci I, Brambilla G, Consonni D, Pesatori AC.
The Seveso studies on early and long-term effects of dioxin
exposure: a review. Environ Health Perspect 1998;106(Suppl
2):625–33.
46. Barnes PJ. Mechanisms in COPD: differences from asthma.
Chest 2000;117(Suppl 2):10S–14S.