Serum Dioxin Levels in Vietnamese Men more than 40 Years afterHerbicide SprayingHo Dung Manh,†,‡ Teruhiko Kido,§,* Rie Okamoto,§ Sun XianLiang,† Le Thai Anh,†

Supratman Supratman,† Shoko Maruzeni,∥ Muneko Nishijo,∥ Hideaki Nakagawa,∥ Seijiro Honma,⊥

Takeshi Nakano,# Takumi Takasuga,▽ Dang Duc Nhu, ▲,○ Nguyen Ngoc Hung, ▲ and Le Ke Son■

†Division of Health Sciences, Graduate School of Medical Science, Kanazawa University, 5-11-80 Kodatsuno, Kanazawa 920-8640,Japan‡Faculty of Environmental Engineering and Biotechnology, Lac Hong University, No.10 Huynh Van Nghe, Buu Long, Bien Hoa,Dong Nai, Vietnam§Faculty of Health Sciences, Institute of Medical Pharmaceutical and Health Sciences, Kanazawa University, 5-11-80 Kodatsuno,Kanazawa 920-8640, Japan∥Department of Public Health, Kanazawa Medical University, 1-1 Daigaku, Uchinada, Ishikawa 920-8641, Japan⊥ASKA Pharma Medical Co. Ltd., 5-36-1 Shimosakunobe, Kawasaki 213-8522, Japan#Center for Advanced Science and Innovation, Osaka University, Osaka 560-0043, Japan▽Shimadzu Techno-Research Inc., Kyoto 604-8435, Japan▲10-80 Division, Hanoi Medical University, No.1 Ton That Tung, Dong Da, Hanoi, Vietnam■Environment Administration, Ministry of Natural Resources and Environment, 67 Nguyen Du Street, Hanoi, Vietnam

ABSTRACT: Recent studies have found elevated dioxin levels inside some U.S.military former air bases in Vietnam, known as hotspots. Many studies of Agent Orangehave been done in U.S. veterans; however, there is little known about Vietnamese men.In 2010, we collected blood samples from 97 men in a hotspot and 85 men in anunsprayed area in Northern Vietnam. Serum concentrations of not only TCDD but alsoother dioxins (PCDDs), furans (PCDFs), and nonortho polychlorinated biphenyls(PCBs) were significantly higher in the hotspot than in the unsprayed area. In thehotspot, three subareas were demarcated, based on their proximity to the air base. Thetotal toxic equivalents (TEQ) of PCDDs/PCDFs+PCBs was 41.7 pg/g lipid in the areaclosest to the air base, while it was around 29 pg/g lipid in the other two subareas. Inthe unsprayed area, the dioxin levels were no different between men who went to theSouth during the Vietnam War and those who remained in the North, with TEQsPCDDs/PCDFs+PCBs of around 13.6 pg/g lipid. Our findings suggested that peopleliving close to the former U.S. air bases might have been exposed to both Agent Orange and other sources of dioxin-likecompounds.

■ INTRODUCTION

Polychlorinated dibenzodioxins (PCDDs), polychlorinateddibenzofurans (PCDFs), and polychlorinated biphenyls(PCBs), all known as dioxins or dioxin-like compounds, arewidespread, persistent toxic chemicals in the environment.Dioxins are lipophilic compounds and accumulate in the humanbody through the food chain. Once in the body, they are storedin adipose tissue and take many years to be eliminated.1

Although the human body burden has been decreasing inrecent years,2 dioxin levels remain elevated in some specificareas where historical incidents occurred.During Operation Ranch Hand (1961−1971), the U.S.

military sprayed millions of liters of various herbicides south ofthe former Demilitarized Zone at the 17th parallel, which wasestablished as a dividing line between North and SouthVietnam as a result of the First Indochina War. A little over

10% of South Vietnam was sprayed for purposes of defoliationand crop destruction. The best-known herbicide was AgentOrange, which is a 1:1 mixture of n-Butyl esters of 2,4-dichlorophenoxyacetic acid (2,4-D) and 2,4,5-trichlorophe-noxyacetic acid (2,4,5-T). Other herbicides were used includingAgent White (2,4-D; picloram), Agent Blue (cacodylic acid),Agent Purple (2,4-D; 2,4,5-T), Agent Green (2,4,5-T), andAgent Pink (2,4,5-T). Unfortunately, 2,4,5-T was contaminatedwith varying levels of 2,3,7,8-tetrachlorodibenzo-p-dioxin(TCDD), the most toxic chemical in the dioxin group.3 As aresult, TCDD in breast milk samples collected in 1970 in the

Received: October 30, 2013Revised: February 15, 2014Accepted: February 19, 2014

Article

pubs.acs.org/est

© XXXX American Chemical Society A dx.doi.org/10.1021/es404853h | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

South showed impressive levels as high as 1832 ppt (parts pertrillion)the highest level in human milk observed to date.4

Two decades later, TCDD levels in pooled blood samples stillproved to be much higher in the South than in the North.5,6

Also, TCDD levels in pooled adipose tissue from Hanoi offormer North Vietnamese soldiers reached as high as 8.1 ppt,compared with 1.4 ppt in the general population pool fromHanoi in North Vietnam.6 Moreover, other PCDD/Fcongeners in adipose tissues7 and sediment samples8 werealso found to be higher in the South than in the North, whichsuggested that part of this difference may be due toatmospheric deposition generated by combustion sources.The Hatfield Consultancy recently theorized that dioxin

levels in former U.S. air bases, where the herbicide was spilledwhen applied by truck-mounted sprayers and also sprayedintensively around the perimeter, remain extremely elevated.9,10

The major air bases at Bien Hoa, Da Nang, and Phu Cat areknown as hotspots.11,12 Tai et al.13 collected breast milk fromyoung mothers who were born after the war and found thattheir dioxin levels were 4-fold higher in the hotspots and 3-foldhigher in sprayed areas compared with unsprayed areas. Otherstudies have found an association between dioxin levels inbreast milk and steroid hormone levels in Vietnameseprimiparae,14−16 as well as an association with the neuro-development17 and body development18 of infants in thehotspot areas.Many studies of Agent Orange have been done in the U.S.

veterans who were exposed for a short time while serving inVietnam.19 Few studies have been conducted on Vietnamesemen who were directly sprayed or lived in herbicide-sprayedareas during the war and may continue to be exposed up tonow. Previous studies quantifying dioxin in the blood ofVietnamese people often used pooled samples;4,6,10 only a fewstudies used individual samples,20 because the analysistechnique required volumes of blood as large as 100 mL,which are difficult to collect in Vietnam. Recently, withadvancements in analytical techniques, we can now analyzedioxins in amounts of blood as small as 10 mL. In 2010, weconducted a study on Vietnamese men located in a hotspot area

versus an unsprayed area. The purpose of this study was todetermine serum dioxin concentrations in these elderly menand compare dioxin levels in the two areas.

■ METHODS

Study Areas. The study areas are shown in Figure 1. PhuCat air base, a dioxin hotspot located in Binh Dinh province inSouth Vietnam, was a former U.S. air base during OperationRanch Hand. Records indicate that 17 000 drums of AgentOrange, 9,000 drums of Agent White, and 2900 drums of AgentBlue were stored at this air base.21 The Hatfield Consultancy9

has reported TCDD levels as high as 236 000 pg/g in soil takenat the herbicide storage area inside the air basefar higher thanthe standard in some countries of less than 1.000 pg-TEQ/g insoil. In the Phu Cat district, we selected three subareas (PC1,PC2, PC3) based on their proximity to the air base. The airbase is inside PC1 (the Cat Tan and Ngo May communes, totalarea 40 km2), near PC2 (Cat Tuong and Cat Trinh communes,total area 80 km2), and a little farther from PC3 (Cat Hanh andCat Lam communes, total area about 110 km2).The comparison site is an unsprayed area in the Kim Bang

district, located in Ha Nam province in the northern part of thecountry. This site was not sprayed with herbicide during theVietnam War. Both the hotspot and the unsprayed area arerural and there is no industrial zone nearby.

Population. Between 2010 and 2011, 97 men in thehotspot area and 85 men in the unsprayed area were recruitedfor this study. All of the subjects were between 55 and 80 yearsof age and agreed to provide 10 mL of blood for testing. Theywere interviewed to obtain their demographic data, includingage, residency, smoking habit, jobs, family income, andeducation. In the unsprayed area, we also asked if they hadany history of exposure to herbicides, whether they had beendirectly sprayed by herbicide or had lived in herbicide-sprayedareas where subsequent defoliation could be recognized, andwhere they had been during the herbicide spraying period(1961−1971), and thereafter. In addition, the height and the

Figure 1. Map of Vietnam and Study Areas.

Environmental Science & Technology Article

dx.doi.org/10.1021/es404853h | Environ. Sci. Technol. XXXX, XXX, XXX−XXXB

weight of each participant were measured to calculate theirbody mass index (BMI, kg/m2).The purpose of the present study was thoroughly explained

to them and written informed consent was obtained from eachparticipant through their local people’s committee. This studywas approved by the Medical Ethics Committee of KanazawaUniversity (Health Permission No. 89).Analysis of Blood Samples. Blood samples (10 mL) were

collected from each participant by medical staff at thecommunity health center. After centrifugation of the samples,the serum was separated into chemically cleaned containers,frozen in dry ice for several days, transported by air toShimadzu Techno-Research Inc., in Kyoto, Japan, and stored at−30 °C until analysis. The analytical technique and qualitycontrol and assurance procedures were implemented followingthe guidelines contained in the publication “Provisional Manualon the Analysis of Dioxin in Human Blood’’ (Japan, Ministry ofHealth and Welfare, 2000).22

The serum samples were spiked with a mixture of 13C12-labelPCDDs/PCDFs and PCBs as an internal standard. The lipidcontent was extracted with n-hexane by liquid extraction andthen determined using the gravimetric method. It was thensubjected to a series of purification operations consisting ofalkaline digestion and multilayer silica gel column chromatog-raphy. Finally, an active carbon-dispersed silica gel column wasused to separate and collect the PCDDs/PCDFs and nonorthoPCBs.The dioxin/furan/nonortho PCB fraction was quantified

using HRGC-HRMS (Hewlett-Packard 6890 Series andMicromass Autospec, Ultima). The selected ion monitoring(SIM) mode was used, and the resolution was maintainedabove 10 000. The spike recovery (calculated using a syringespike) ranged between 95% and 104% for PCBs, 78% and 89%for PCDD/Fs.23

All dioxin, furan, and nonortho PCB congeners werecalculated on a lipid basis, then converted to toxic equivalents(TEQs) using the international toxicity equivalency factors(TEFs) 2005 recommended by the World Health Organization(WHO).24 The detection limits of congeners varied dependingon the samples. Generally, a serum sample of 5 g wet weightwith 0.5% fat content (w/w) has detection limits, as follows:Te-PeCDD/Fs: 0.01 pg/g wet, 2 pg/g fat; Hx-HpCDD/Fs:0.02 pg/g wet, 4 pg/g fat; OCDD/Fs: 0.05 pg/g wet, 10 pg/gfat; and PCB #81, #77, #126, #169: 0.1 pg/g wet, 20 pg/g fat.Serum concentrations of dioxin, furan, and nonortho PCBcongeners below the limit of detection (LOD) were assigned avalue equal to half the LOD.Statistical Analysis. We transformed the dioxin concen-

trations into log10 to improve normality. Student’s t test and thechi-square test were used to compare the two groups accordingto continuous or categorical variables. For more than twogroups, ANOVA was used first and then Tukey’s posthoc testto find which pair was significantly different. We used multiplelinear regressions to compare dioxin concentration levelsbetween the areas after adjusting for age, BMI, smoking, andresidency. Since the serum concentration used in the regressionmodel was transformed into log10, the regression coefficient waspresented as 10β, where β is the regression coefficient. For thecategorical variables, this number describes the fold change inthe dioxin concentration (an increase if 10β > 1, or a decrease if10β < 1) for each category compared with the referencecategory.

All statistical analyses were performed using the JMP@9software package (SAS Institute, Japan) and the R StatisticalEnvironment (R Development Core Team, 2013).

■ RESULTSCharacteristics of the Participants. Table 1 shows the

demographic characteristics of the participants in the two areas.

The average age of the men in the hotspot was 68 years, similarto the average age of 67 years for the men in the unsprayedarea. The participants’ height and the weight were higher in theunsprayed area, but the BMI was not different. The percentageof former soldiers in the unsprayed area was higher than in thehotspot. The rates of smoking and alcohol use were similar forthe two groups. In the unsprayed area, 67% of the men had aneducation level above secondary school, while that rate was48% in the hotspot area. The percentage of farmers in thehotspot area was 55% and in the unsprayed area, 32%.

Dioxin Levels in the Hotspot and the Unsprayed Area.Table 2 compares dioxin levels in the hot spot with those in theunsprayed area. Most of the dioxin, furan, and nonortho PCBconcentrations were higher in the hotspot than in theunsprayed area. The geometric mean TCDD level in thehotspot was 2.6 pg/g lipid, while it was 1.5 pg/g lipid in theunsprayed area. The total TEQs of PCDD/Fs + PCBs in thehotspot was about 2.5-fold higher than in the unsprayed area.Figure 2 shows TCDD levels in the three subareas of the

hotspot and the unsprayed area. The TCDD levels in PC1 werethe highest in all the areas. None of the men in the KB hadTCDD levels higher than 5 pg/g lipid, while 11 men (28%) inPC1, and 2 men (7%) in PC2, and 1 man (3%) in PC3 hadTCDD levels higher than 5 pg/g lipid. In PC1, two of the menwho had worked inside the air base had the highestconcentrations of TCDD (16 and 24 pg/g lipid, respective-ly).These results suggest that highly exposed men are prevalent,particularly in PC1the closest area to the air base.

Table 1. Demographic Characteristics of the StudyParticipantsa

hotspot (n = 97) unsprayed (n = 85)

continuous variables mean SD mean SD p value

age (years) 68 6 67 5 0.9height (cm) 157 5 159 5 0.001weight (kg) 49 7 52 8 0.005BMI (kg/m2) 20 2 21 3 0.1residency (years) 65 11 58 10 0.001income (*1000 VND)a 1300 1380 1314 1246 0.4categorical variables N % N %former soldier 33 34 67 79 0.0001smoking 60 62 62 73 0.1alcohol 46 47 42 49 0.8education

<secondary school 50 52 20 24 0.0001≥secondary school 47 48 65 76

currently employed (yes) 67 69 43 51 0.01farmer 53 55 27 32 0.01laborer 0 0 2 2fisherman 0 0 1 1teacher 0 0 1 1other 14 14 12 14

amean ± SD or N (%).

Environmental Science & Technology Article

dx.doi.org/10.1021/es404853h | Environ. Sci. Technol. XXXX, XXX, XXX−XXXC

More than 70% of the 2378-TeCDF, 12378-PeCDF, 123789-HxCDF, 234678-HxCDF, 1234789-HpCDF, OCDF, TeCB#81, and TeCB #77 congeners were below detection limits ornondetected (ND) in both areas. Since their contribution tototal TEQ dioxins was small, they are not shown in all of thetables.Table 3 gives the dioxin levels for the three subareas in the

hotspot. Some PCDD congeners and nonortho PCB congeners(PCB126 and PCB169) were significantly higher in PC1 thanin PC2 or PC3. The PCDF congener levels were no differentbetween the three subareas except for 23478-PeCDF. None ofthe congener levels were significantly different between PC2and PC3.Table 4 shows the dioxin levels for the three groups in the

unsprayed area. Group A consisted of men who had remainedin the North their entire lifetime. Group B were men whostayed in the South at some time during the herbicide spraying(1961−1971), with an average stay of 4.5 years. Group C weremen who went to the South at some time after 1971, with anaverage stay of 7.8 years. There was no significant difference indioxin levels between the three groups, even after using

multiple linear regressions to adjust for age, BMI, residency,and smoking. In Group B, we asked whether they had everbeen directly sprayed by an herbicide during 1961−1971 or hadever lived in sprayed areas where subsequent defoliation couldbe recognized. We compared dioxin levels between 5 men whoanswered “no” to both questions and 33 men who answered“yes” to at least one question, but we found no significantdifference.Table 5 shows dioxin concentrations in the three subareas of

the hotspot relative to the unsprayed area (KB) after adjustingfor age, BMI, smoking, and residency. β is the regressioncoefficient, and 10β tells how many times higher the dioxinconcentrations were in each subarea of the hotspot comparedwith the unsprayed area. The TEQ of PCDD/Fs + PCBs wasabout 3-fold higher in PC1 and 2-fold higher in PC2 and PC3than it was in KB. According to the model, the percentage ofvariation (R2) for the total TEQ of PCDD/Fs + PCBs was 0.52.

■ DISCUSSIONTo our knowledge, this is one of a few studies to examine theindividual serum dioxin levels in Vietnamese men whoexperienced herbicide spraying at any time during the VietnamWar. We found that serum dioxin levels were higher in thehotspot than in the unsprayed area. In the hotspot, dioxinTEQs in men living close to the former air base were 3-foldhigher, and in men living near and a little farther from theairbase they were 2-fold higher compared with the unsprayedarea. In contrast, serum dioxin levels were not significantlydifferent between the men in the North who went to the Southduring the Vietnam War and those who remained in the North.TCDD is known to be the marker for Agent Orange, with

levels as high as 1832 ppt detected in breast milk samplesshortly after exposure.4 Although the dioxin levels decrease overtime, high TCDD levels were still found as high as 413 ppt inblood samples and in food such as duck, chicken, and fishcollected during 1991−2000 in Bien Hoa city.20,25 Thesestudies suggested that Bien Hoa Airbase, where Agent Orangewas stored for spraying missions, was the origin of TCDD.Dwernychuk et al.10 studied in the Aluoi Valley, in central

Table 2. Dioxin Concentrations in the Hotspot and the Unsprayed Area

Phu Cat (n = 97) Kim Bang (n = 85)

Pg/g lipid geomean median (interquartile) geomean median (interquartile) p value

2378-TeCDD 2.6 2.6 (1.6−3.7) 1.5 1.5 (1.2−2.0) ***12378-PeCDD 8.3 8.3 (5.5−13) 2.4 2.4 (1.5−3.5) ***123478-HxCDD 5.1 4.8 (3.4−7.0) 2.7 2.9 (2.0−3.9) ***123678-HxCDD 19 19 (14−26) 4.5 4.3 (3.0−5.7) ***123789-HxCDD 6.5 6.5 (4.1−9.8) 2.8 2.9 (2.0−3.9) ***1234678-HpCDD 29 27 (17−44) 6.6 6.2 (4.7−9.6) ***OCDD 320 310 (200−505) 59 57 (38−96) ***23478-PeCDF 13 14 (9−18) 7.1 6.8 (5.2−9.5) ***123478-HxCDF 28 26 (18−41) 4.3 4.5 (2.9−5.7) ***123678-HxCDF 21 21 (13−32) 4.9 4.8 (3.2−6.3) ***1234678-HpCDF 41 42 (26−65) 4.0 3.9 (2.7−5.8) ***PeCB #126 36 31 (19−56) 30 29 (16−54) 0.1HxCB #169 114 110 (72−160) 24 23 (15−39) ***TEQ pg/g lipidtotal PCDDs 14.9 14.1 (9.8−21.5) 5.2 5.2 (3.8−6.7) ***total PCDFs 10.5 10.0 (7.2−14.9) 4.0 3.8 (2.9−5.1) ***total PCDD/Fs 26 25 (18−36) 9.2 8.9 (7.2−11.6) ***nonortho PCBs 7.3 6.4 (4.7−11) 3.9 3.6 (2.4−6.3) ***total PCDD/Fs + PCBs 33.6 30.9 (22.7−46.9) 13.6 12.9 (9.8−17.5) ***

Figure 2. TCDD levels in the hotspot and the nonexposed area.

Environmental Science & Technology Article

dx.doi.org/10.1021/es404853h | Environ. Sci. Technol. XXXX, XXX, XXX−XXXD

Vietnam, and found that TCDD levels were as high as 41 pptand contributed to more than 80% of the total TEQ PCDD/Fsin human blood samples collected near former Aso air base. Inthe present study, we found that the dioxin levels increasedwith proximity to the Phu Cat air base. The geometric means ofthe TCDD levels in PC1, PC2, PC3, and KB were 3.5, 2.4, 2.0,and 1.5 pg/g lipid, respectively. However, these TCDD levelswere low and only contributed to about 10% total TEQPCDD/Fs, which is similar to the TCDD level found in breastmilk in a recent study.13 The reason for the low TCDD levelmay be due to time-dependent changes in the pattern of dioxinresidue when passed through many livings organism after fourdecades.

Dioxins come from many sources, which have specificpatterns for PCDD/F congeners; TCDD was the maincontamination in Agent Orange. The present study foundelevation of not only TCDD but also other higher chlorinatedPCDDs, PCDFs, and PCB169 in the hotspot area, whichsuggests there may be other dioxin sources besides theherbicide used in the past, such as incineration or activitiesoccurring at the air base. Tawara et al.26 found that breast milkin a sprayed area contained elevated levels of chlorinated dioxincompounds and hypothesized that there could be PCPcontamination from tactical herbicides sprayed in Vietnam.Recent studies found potential dioxin sources in Vietnam, suchas open municipal dumping sites27 and e-waste recycling sites.28

Table 3. Dioxin Concentrations in the Three Sub-Areas of the Hotspot

PC1, n = 39 PC2, n = 27 PC3, n = 31 p valuea

Pg/g lipidgeo-mean

median(interquartile)

geo-mean

median(interquartile)

geo-mean

median(interquartile)

PC1-PC2

PC1-PC3

2378-TeCDD 3.5 3.3 (1.9−5.6) 2.4 2.5 (1.8−3.4) 2.0 2.2 (1.4-.2.9) ***12378-PeCDD 10.8 10 (7.4−16) 7.4 7 (5−11) 6.7 6.6 (4.2−12) * **123478-HxCDD 6.1 5.8 (3.7−8.3) 4.5 4.3 (3.5−5.7) 4.5 4.5 (2.9−6.1)123678-HxCDD 24.1 22 (18−38) 15.1 15 (11−20) 17.0 16 (11−26) ** *123789-HxCDD 7.4 6.9 (5−10) 5.6 6.5 (3.6−7.3) 6.2 6 (3.6−11)1234678-HpCDD 33.4 32 (22−54) 21.2 21 (15−33) 30.1 28 (18−44) *OCDD 370 340 (230−540) 294 300 (200−450) 286 270 (160−410)23478-PeCDF 15.5 15 (12−21) 12.2 12 (9.8−14) 11.5 11 (6.7−17) *123478-HxCDF 29.2 27 (20−44) 24.8 26 (16−34) 28.6 24 (16−52)123678-HxCDF 22.3 22 (14−34) 19.0 20 (13−26) 21.6 20 (13−40)1234678-HpCDF 41.6 42 (29−64) 36.1 38 (20−58) 45.0 42 (26−88)PeCB #126 52.8 44 (31−83) 25.7 24 (17.5−34) 28.5 27 (16.5−49) *** ***HxCB #169 147 150 (86−220) 100 100 (68−130) 92 89 (66−140) *** *TEQ pg/g lipidtotal PCDDs 19.0 19.3 (12.8−27.7) 13.1 11.8 (9.7−17.5) 12.3 12.6 (8.4−19.9) ** *total PCDFs 11.5 10.9 (7.9−15.6) 9.6 9.6 (7.4−11.4) 10.1 9.6 (6.8−17.7)total PCDD/Fs 31.0 29.8 (23.7−41.2) 23.0 23.2 (17.1−27.5) 22.9 19.9 (16.1−36.2) * *total nonortho PCBs 10.2 9.9 (5.8−14.1) 5.8 5.5 (4.2−6.5) 5.8 5.4 (4.3−5.4) *** ***total PCDD/Fs + nonortho PCBs 41.7 41.7 (28.8−53.8) 29.2 28.6 (21.4−34.7) 28.9 24.9 (22.2−45.8) ** **a*, p < 0.05; **, p < 0.01; and ***, p < 0.001.

Table 4. Dioxin Concentrations in the Three Groups in the Unsprayed Areaa

group A (n = 35) group B (n = 38) group C (n = 12)

Pg/g lipid geo-mean median (interquartile) geo-mean median (interquartile) geo-mean median (interquartile)

2378-TeCDD 1.5 1.5 (1.2−2) 1.4 1.4 (1.2−1.7) 1.6 1.6 (1−2.2)12378-PeCDD 2.5 2.4 (1.5−4) 2.2 2.4 (1.5−2.9) 2.7 2.7 (1.5−4.6)123678-HxCDD 4.5 4.1 (2.9−5.6) 4.3 4.5 (3−5.7) 4.8 4.6 (3.2−8.2)1234678-HpCDD 6.8 5.8 (4.8−11) 6.5 6.6 (4.8−9.2) 6.1 6.6 (4.4−9.3)OCDD 56.5 47 (33−96) 57.1 57 (36−88.8) 76 81 (45−138)23478-PeCDF 7.4 7.1 (4.8−11) 6.6 6 (5.2−8.4) 8.2 7.4 (5.8−13)123478-HxCDF 4.2 3.9 (2.8−5.6) 4.2 4.5 (2.9−5.5) 5.5 5.5 (3.2−9.7)123678-HxCDF 4.9 5.3 (3.2−6.4) 4.6 4.7 (3.3−5.7) 5.5 5 (3.2−12.1)1234678-HpCDF 3.7 3.7 (2.7−5.8) 4.0 4.1 (2.7−5.6) 5.1 5.1 (2.9−9.3)PeCB #126 33.3 31 (18.5−57) 28.7 28 (15−51) 27.4 27.5 (16.8−43.4)HxCB #169 26.3 23.5 (16−40) 21.2 22.8 (13.9−28.3) 24.5 21.5 (15.3−43)TEQ pg/g lipidtotal PCDDs 5.4 5.4 (3.6−7.8) 4.8 4.8 (4−6.1) 5.8 5.3 (3.9−8.3)total PCDFs 4.0 4.2 (2.6−5.2) 3.7 3.7 (3−4.4) 4.6 4 (3.5−7.6)total PCDD/Fs 9.6 9.7 (6.4−13.1) 8.6 8.6 (7.3−10.4) 10.4 10.1 (7.6−15.2)total nonortho PCBs 4.4 4 (2.6−7.3) 3.6 3.5 (2−5.8) 3.6 3.6 (2.6−6.2)total PCDD/Fs + nonortho PCBs 14.5 13.5 (9.1−21.4) 12.6 12.5 (10.2−16) 14.2 13.8 (9.8−19.6)

aGroup A: men who remained in the North their entire life; Group B: men who went to the South at some time during the herbicide spraying(1961−1971); Group C: men who went to the South at any time after 1971.

Environmental Science & Technology Article

dx.doi.org/10.1021/es404853h | Environ. Sci. Technol. XXXX, XXX, XXX−XXXE

Other studies also found organohalogen compounds, whichmay come from industrial sources and pesticides used inagriculture.29 Although we cannot identify the contaminationsource of dioxins other than TCDD in the herbicide, ourevidence justifies further considerationIn the unsprayed area, dioxin levels were not found to be

different from those in men who went to the South during theVietnam War and those who remained in the North. Since nostudy has been done on the half-life of dioxins in Vietnamesemen, we assumed a half-life for TCDD of seven years based onstudies of U.S. veterans.1 Schecter et al.6 reported that theTCDD level in the adipose tissue of former soldiers from theNorth collected in 1987−1991 was about 8.1 ppt. Since aboutthree half-lives have elapsed since then, if the men were notexposed to dioxin after returning to the North, their levelshould be about 1 ppt, which is close to the average of 1.5 pptfound in serum in the present study. Actually, not allVietnamese soldiers who went to the South were exposed toherbicide; some may have no or little exposure. The results inthe Phu Cat hotspot showed that former soldiers who workedinside the air base had the highest TCDD levels. A study of theAir Force Ranch Hand personnel who were responsible for theaerial spraying of herbicides from fixed-wing aircraft in Vietnambetween 1962 and 1971 showed that depending on the jobtype, serum TCDD levels collected in 1987 from the officers,pilots, and enlisted ground crew had median levels of 8, 18, and24 ppt, respectively.30 Moreover, serum samples collected in1999−2000 from members of the Army Chemical Corps whowere responsible for the spraying of herbicides around theperimeters of base camps and for aerial spraying fromhelicopters in Vietnam yielded TCDD levels of 3−4 ppt.31

By now, the dioxin levels in both the North Vietnameseveterans and the U.S. veterans who were involved in the RanchHand Operation have decreased significantly and are nowprobably as low as those in the general population.

Some limitations should be considered in this study. First,the mens’ history of herbicide exposure was self-reported. Ittook time and effort for these elderly men to recall the events.Also, we did not find any correlation between herbicideexposure and dioxin levels in Northern men who went to theSouth during the Vietnam War, probably because they were notexposed to dioxins for a long time after returning to the North.In the hotspot, we could not indentify contaminated sourcesbesides the herbicide used in the past. In future studies, moreinformation should be collected on food intake in order tobetter understand the serum dioxin levels in these elderly men,especially the ones who received the most exposure in PC1could be useful for learning about exposure pathways. Also,recent studies of breast milk suggested that food intake maycontribute only slightly to the increased dioxin levels in the DaNang hotspot,32 and past exposure rather than present dietaryintake may affect present dioxin levels in the sprayed area33 inVietnam.For public health, hotspots are the areas of concern with

regard to dioxin exposure. Some public heath interventionprograms have been conducted to reduce the risk of dioxinexposure through food for local residents living around the airbases.34,35 In addition, dioxin remediation activities supportedby the United States and Vietnamese governments are currentlyunder way inside the hotspots.36 Still, it is very important tomonitor dioxins levels regularly and study any adverse healtheffects on people living in or near these hotspots to confirm thebenefits of the programs.This study analyzing individual serum dioxin levels in

Vietnamese men in a hotspot and an unsprayed area foundthat dioxin levels were higher in the hotspot area than in theunsprayed area. In the hotspot area, the dioxin levels increasedwith proximity to the air base. In the unsprayed area, however,dioxin levels were not elevated in the men from the North whowent to the South during the Vietnam War compared with the

Table 5. Dioxin Concentrations in Three Areas of the Hotspot Compared with the Unsprayed Area, Adjusted for Age, BMI,Smoking, And Residencya

PC1 PC2 PC3 KB

Pg/g lipid 10β p 95% CIb adj. GMc 10β p 95% CIb adj. GMc 10β p 95% CIb adj. GMc adj. GMc R2

2378-TeCDD 2.3 *** 1.9−2.9 3.6 1.6 1.2−2..0 2.4 1.2 0.9−1.6 1.9 1.5 0.3112378-PeCDD 4.4 *** 3.4−5.6 11.2 3.0 * 2.2−4.0 7.6 2.5 1.9−3.4 6.5 2.5 0.52123478-HxCDD 2.2 *** 1.8−2.7 6.3 1.6 1.3−2.0 4.6 1.5 1.2−1.9 4.3 2.9 0.33123678-HxCDD 5.3 *** 4.2−6.5 24.8 3.3 * 2.6−4.3 15.7 3.5 ** 2.8−4.6 16.7 4.7 0.65123789-HxCDD 2.5 *** 2.0−3.1 7.6 1.8 1.4−2.4 5.5 1.9 1.5−2.4 5.7 3.0 0.381234678-HpCDD 4.9 *** 3.9−6.1 34.7 3.2 2.5−4.1 22.7 4.3 *** 3.3−5.5 30.2 7.1 0.65OCDD 6.1 *** 4.9−7.7 379 5.0 ** 3.8−6.6 311 4.7 3.6−6.1 290 61.8 0.6923478-PeCDF 2.1 *** 1.7−2.6 15.8 1.6 1.3−2.0 12.3 1.5 1.2−1.8 11.0 7.5 0.32123478-HxCDF 6.5 *** 5.2−8.0 29.9 5.4 *** 4.2−7.0 25.2 5.9 *** 4.6−7.6 27.3 4.6 0.75123678-HxCDF 4.4 *** 3.5−5.4 23.1 3.7 ** 2.9−4.7 19.4 3.9 *** 3.0−5.0 20.4 5.3 0.661234678-HpCDF 9.8 *** 7.6−13 42.8 8.2 *** 6.1−11 36.0 9.6 *** 7.1−13 41.8 4.4 0.76PeCB #126 1.8 *** 1.3−2.3 53.7 0.9 0.6−1.2 27.6 1.0 0.7−1.3 29.7 30.6 0.15HxCB #169 6.0 *** 4.7−7.6 152 4.0 ** 3.0−5.2 101 3.4 2.6−4.5 85.8 25.4 0.65TEQ pg/g lipidtotal PCDDs 3.5 *** 2.9−4.3 19.6 2.4 1.9−3.0 13.3 2.1 1.7−2.7 11.8 5.5 0.55total PCDFs 2.8 *** 2.3−3.3 11.8 2.3 * 1.9−2.8 9.7 2.3 * 1.9−2.8 9.6 4.2 0.56total PCDD/Fs 3.2 *** 2.7−3.9 31.9 2.4 * 1.9−2.9 23.3 2.2 1.8−2.7 21.9 9.8 0.58total nonorthor PCBs 2.6 *** 2.0−3.3 10.5 1.5 1.1−2.0 6.1 1.4 1.1−1.9 5.8 4.0 0.30total PCDD/Fs +nonorthor PCBs

3.0 *** 2.5−3.6 42.9 2.1 1.7−2.6 30.0 2.0 1.6−2.4 28.1 14.4 0.52

aβ is the regression coefficient based on log10 (dioxin concentration)b95% CI: confidence interval of 10β cAdj. GM: geometric mean adjusted by age,

BMI, smoking and residency

Environmental Science & Technology Article

dx.doi.org/10.1021/es404853h | Environ. Sci. Technol. XXXX, XXX, XXX−XXXF

men who remained in the North. Therefore, public heathprevention and dioxin monitoring programs should focus inpeople living in the areas surrounding these hotspots.

■ AUTHOR INFORMATION

Corresponding Author*Tel: +81-76-265-2565; fax+81-76-265-2565; e-mail: kido@mhs.mp.kanazawa-u.ac.jp.

Present Address○School of Medicine and Pharmacy, Vietnam NationalUniversity, 144 Xuan Thuy, Cau Giay, Hanoi, Vietnam

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

The study was supported by a Grant-in-Aid for ScientificResearch (B) from the Japan Society for the Promotion ofScience, No. 23406018.

■ REFERENCES(1) Milbrath, M. O.; Wenger, Y.; Chang, C. W.; Emond, C.;Garabrant, D.; Gillespie, B. W.; Jolliet, O. Apparent half-lives ofdioxins, furans, and polychlorinated biphenyls as a function of age,body fat, smoking status, and breast-feeding. Environ. Health Perspect.2009, 117, 417−425.(2) Furst, P. Dioxins, polychlorinated biphenyls and other organo-halogen compounds in human milk. Levels, correlations, trends andexposure through breastfeeding. Mol. Nutr. Food Res. 2006, 50, 922−933.(3) Stellman, J. M.; Stellman, S. D.; Christian, R.; Weber, T.;Tomasallo, C. The extent and patterns of usage of Agent Orange andother herbicides in Vietnam. Nature 2003, 422, 681−687.(4) Schecter, A.; Dai, L. C.; Thuy, L. T.; Quynh, H. T.; Minh, D. Q.;Cau, H. D.; Phiet, P. H.; Nguyen, N. T.; Constable, J. D.; Baughman,R.; et al. Agent Orange and the Vietnamese: The persistence ofelevated dioxin levels in human tissues. Am. J. Public Health 1995, 85,516−522.(5) Schecter, A.; Mcgee, H.; Stanley, J.; Boggess, K. Dioxin,dibenzofuran, and PCB, including coplanar PCB levels in the bloodof Vietnam veterans in the Michigan Agent Orange study. Chemosphere1992, 25, 205−208.(6) Schecter, A.; Pap̈ke, O.; Ball, M.; Cau, H. D.; Dai, L. C.; Ming, N.Q.; Quynh, H. T.; Phuong, N. N. T.; Phiet, P. H.; Chi, H. K.; et al.Dioxin and dibenzofuran levels in blood and adipose tissue ofVietnamese from various locations in Vietnam in proximity to AgentOrange Spraying. Chemosphere 1992, 25, 1123−1128.(7) Schecter, A.; Ryan, J. J.; Constable, J. D. Chlorinated dibenzo-p-dioxin and dibenzofuran levels in human adipose tissue and milksamples from the north and south of Vietnam. Chemosphere 1986, 15,1613−1620.(8) Schecter, A.; Eitzer, B. D.; Hites, R. A. Chlorinated dioxin anddibenzofuran levels in sediments collected from rivers in Vietnam.Chemosphere 1989, 18, 831−834.(9) Vietnam-Russia Tropical Center and Hatfield Consultants.Evaluation of Contamination at the Agent Orange dioxin hotspotsin Bien Hoa, Phu Cat and Vicinity, Viet Nam. Final Report, June 2009.h t t p : / /www . h a t fi e l d g r o up . c om/wp - c on t e n t / u p l o a d s /AgentOrangeReports/UNDP1391/UNDP1391_Final_VRT_%20Report_and_Appendices.pdf.(10) Dwernychuk, L. W.; Cau, H. D.; Hatfield, C. T.; Boivin, T. G.;Hung, T. M.; Dung, P. T.; Thai, N. D. Dioxin reservoirs in southernVietnamA legacy of Agent Orange. Chemosphere 2002, 47, 117−137.(11) Dwernychuk, L. W. Dioxin hot spots in Vietnam. Chemosphere2005, 60, 998−999.

(12) Dwernychuk, L. W.; Hung, T. M.; Boivin, T. G.; Bruce, G. S.;Dung, P. T.; Son, L. K.; Hatfield, C. T.; Dung, N. T.; Allan, J. A.; Nhu,D. D.; et al. The Agent Orange dioxin issue in Vietnam: A manageableproblem. Organohalogen compounds 2006, 312−315.(13) Tai, P. T.; Nishijo, M.; Kido, T.; Nakagawa, H.; Maruzeni, S.;Naganuma, R.; Anh, N. T.; Morikawa, Y.; Luong, H. V.; Anh, T. H.;et al. Dioxin concentrations in breast milk of Vietnamese nursingmothers: A survey four decades after the herbicide spraying. Environ.Sci. Technol. 2011, 45, 6625−6632.(14) Nhu, D. D.; Kido, T.; Naganuma, R.; Suzuki, H.; Kuroda, N.;Honma, S.; Tai, P. T.; Maruzeni, S.; Nishijo, M.; Nakagawa, H.; et al.Salivary cortisol and cortisone levels, and breast milk dioxinconcentrations in Vietnamese primiparas. Toxicol. Environ. Chem.2010, 92, 1939−1952.(15) Manh, H. D.; Kido, T.; Okamoto, R.; Sun, X. L.; Viet, N. H.;Nakano, M.; Tai, P. T.; Maruzeni, S.; Nishijo, M.; Nakagawa, H.; et al.The relationship between dioxins and salivary steroid hormones inVietnamese primiparae. Environ. Health Prev. Med. 2013, 18, 221−229.(16) Kido, T.; Dao, T. V.; Ho, M. D.; Dang, N. D.; Pham, N. T.;Okamoto, R.; Pham, T. T.; Maruzeni, S.; Nishijo, M.; Nakagawa, H.;et al. High cortisol and cortisone levels are associated with breast milkdioxin concentrations in Vietnamese women. Eur. J. Endocrinol. 2013,170, 131−139.(17) Tai, P. T.; Nishijo, M.; Anh, N. T.; Maruzeni, S.; Nakagawa, H.;Luong, H. V.; Anh, T. H.; Honda, R.; Kido, T.; Nishijo, H. Dioxinexposure in breast milk and infant neurodevelopment in Vietnam.Occup. Environ. Med. 2013, 70, 656−62.(18) Nishijo, M.; Tai, P. T.; Nakagawa, H.; Maruzeni, S.; Anh, N. T.;Luong, H. V.; Anh, T. H.; Honda, R.; Morikawa, Y.; Kido, T.; et al.Impact of perinatal dioxin exposure on infant growth: a cross-sectionaland longitudinal studies in dioxin-contaminated areas in Vietnam. PloSone 2012, 7, e40273.(19) National Research Council. Veterans and Agent Orange: Update2008; Washington, DC: The National Academies Press, 2009.(20) Schecter, A.; Dai, L. C.; Pap̈ke, O.; Prange, J.; Constable, J. D.;Matsuda, M.; Thao, V. D.; Piskac, A. L. Recent dioxin contaminationfrom Agent Orange in residents of a southern Vietnam city. J. Occup.Environ. Med. 2001, 43, 435−43.(21) Young, A. L. The History, Use, Disposition and Environmental Fateof Agent Orange; Springer-Verlag: New York, 2009.(22) Ministry of Health and Welfare. Japan, 2000. Provisional Manualon the Analysis of Dioxin in Human Blood, Tokyo (in Japanese).(23) Takasuga, T.; Senthilkumar, K.; Takemori, H.; Ohi, E.; Tsuji,H.; Nagayama, J. Impact of FEBRA (fermented brown rice withAspergillus oryzae) intake and concentrations of PCDDs, PCDFs andPCBs in blood of humans from Japan. Chemosphere 2004, 57, 1409−1426.(24) Van den Berg, M.; Birnbaum, L. S.; Denison, M.; De Vito, M.;Farland, W.; Feeley, M.; Fiedler, H.; Hakansson, H.; Hanberg, A.;Haws, L.; et al. The 2005 World Health Organization reevaluation ofhuman and mammalian toxic equivalency factors for dioxins anddioxin-like compounds. Toxicol. Sci. 2006, 93, 223−241.(25) Schecter, A.; Quynh, H. T.; Pavuk, M.; Papke, O.; Malisch, R.;Constable, J. D. Food as a source of dioxin exposure in the residents ofBien Hoa City, Vietnam. J. Occup. Environ. Med. 2003, 45, 781−788.(26) Tawara, K.; Nishijo, M.; Maruzeni, S.; Nakagawa, H.; Kido, T.;Naganuma, R.; Suzuki, H.; Nhu, D. D.; Hung, N. N.; le Thom, T. H.Residual congener pattern of dioxins in human breast milk in southernVietnam. Chemosphere 2011, 84, 979−986.(27) Nguyen, H. M.; Tu, B. M.; Watanabe, M.; Kunisue, T.;Monirith, I.; Tanabe, S.; Sakai, S.; Subramanian, A.; Sasikumar, K.;Pham, H. V.; et al. Open dumping site in Asian developing countries: apotential source of polychlorinated dibenz-p-dioxins and polychlori-nated dibenzofurans. Environ. Sci. Technol. 2003, 37, 1493−1502.(28) Tue, N. M.; Suzuki, G.; Takahashi, S.; Isobe, T.; Trang, P. T.;Viet, P. H.; Tanabe, S. Evaluation of dioxin-like activities in settledhouse dust from Vietnamese e-waste recycling sites: relevance ofpolychlorinated/brominated dibenzo-p-dioxin/furans and dioxin-likePCBs. Environ. Sci. Technol. 2010, 44, 9195−9200.

Environmental Science & Technology Article

dx.doi.org/10.1021/es404853h | Environ. Sci. Technol. XXXX, XXX, XXX−XXXG

(29) Minh, N. H.; Minh, T. B.; Iwata, H.; Kajiwara, N.; Kunisue, T.;Takahashi, S.; Viet, P. H.; Tuyen, B. C.; Tanabe, S. Persistent organicpollutants in sediments from Sai Gon-Dong Nai River basin, Vietnam:levels and temporal trends. Arch. Environ. Contam. Toxicol. 2007, 52,458−465.(30) Buffler, P. A.; Ginevan, M. E.; Mandel, J. S.; Watkins, D. K. TheAir Force health study: An epidemiologic retrospective. Ann.Epidemiol. 2011, 21, 673−687.(31) Kang, H. K.; Dalager, N. A.; Needham, L. L.; Patterson, D. G.,Jr.; Lees, P. S.; Yates, K.; Matanoski, G. M. Health status of ArmyChemical Corps Vietnam veterans who sprayed defoliant in Vietnam.Am. J. Ind. Med. 2006, 49, 875−884.(32) Anh, N. T.; Nishijo, M.; Tai, P. T.; Maruzeni, S.; Morikawa, Y.;Anh, T. H.; Luong, H. V.; Dam, P. M.,; Nakagawa, H.; Son, L. K. et al.Maternal risk factors associated with increased dioxin concentrationsin breast milk in a hot spot of dioxin contamination in Vietnam. J.Expo. Sci. Environ. Epidemiol.. 2013 Oct 23. [Epub ahead of print](33) Saito, K.; Nhu, D. D.; Suzuki, H.; Kido, T.; Naganuma, R.;Sakakibara, C.; Tawara, K.; Nishijo, M.; Nakagawa, H.; Kusama, K.;et al. Association between dioxin concentrations in breast milk andfood group intake in Vietnam. Environ. Health Prev. Med. 2010, 15,48−56.(34) Tran, T. T. H.; Le, V. A.; Nguyen, N. B.; Tenkate, T.Environmental health risk assessment of dioxin exposure throughfoods in a dioxin hot spotBien Hoa City, Vietnam. Int. J. Environ.Res. Public Health 2010, 7, 2395−2406.(35) Tran, T. H.; Nguyen, N. B.; Le, V. A. New strategy towarddioxin risk reduction for local residents surrounding severe dioxinhotspots in Vietnam. Global Health Action 2013, 6, 21105.(36) Young, A. L.; Van Houten, W. J.; Andrews, W. B., 2nd AgentOrange and Dioxin Remediation Workshop. Hanoi, Vietnam, 18−20June 2007. Environ. Sci. Pollut. Res. Int. 2008, 15, 113−118.

Environmental Science & Technology Article

dx.doi.org/10.1021/es404853h | Environ. Sci. Technol. XXXX, XXX, XXX−XXXH