Based on new evidence and a review of prior studies, the current committee did not find any new associations between exposure to the chemicals of interest (COIs; 2,4-dichlorophenoxyacetic acid [2,4-D], 2,4,5-trichlorophenoxyacetic acid [2,4,5-T], picloram, dimethylarsinic acid [DMA or cacodylic acid], and 2,3,7,8-tetrachlorodibenzo-p-dioxin [TCDD]) and other chronic, non-malignant outcomes including respiratory disorders, gastrointestinal and digestive diseases, adverse effects on thyroid homeostasis, kidney and urinary disorders, chronic skin conditions, eye problems, and bone conditions. Current evidence supports the results of earlier updates that:
- There is sufficient evidence of an association between the COIs and chloracne.
- There is limited or suggestive evidence of an association between exposure to the COIs and hypothyroidism, early onset peripheral neuropathy, and porphyria cutanea tarda.
- There is inadequate or insufficient evidence to determine whether there is an association between exposure to the COIs and respiratory disorders, gastrointestinal and digestive disease (including liver toxicity), kidney and urinary disorders, adverse effects on endocrine function (other than hypothyroidism), chronic skin conditions, eye problems, or bone conditions.
This chapter discusses data on the possible association between exposure to the COIs and several non-cancer health outcomes: respiratory disorders, gastrointestinal and digestive disease (including liver toxicity), adverse effects on thyroid
homeostasis, kidney and urinary disorders, chronic skin conditions, eye problems, and bone conditions. The committee also considers the results of studies of exposure to polychlorinated biphenyls (PCBs) and other dioxin-like chemicals to be informative if they were reported in terms of TCDD toxic equivalents (TEQs) or as concentrations of dioxin-like specific congeners. Although all studies reporting TEQs based on PCBs were reviewed, TEQs based only on mono-ortho PCBs (which are PCBs 105, 114, 118, 123, 156, 157, 167, and 189) are several orders of magnitude lower than those of the non-ortho PCBs (77, 81, 126, and 169), based on the revised World Health Organization (WHO) toxicity equivalency factor (TEF) scheme of 2005 (La Rocca et al., 2008; van den Berg et al., 2006). The lower TEQs of the mono-ortho PCBs, however, may be counterbalanced by their abundance, which is generally many orders of magnitude higher than that of the non-ortho PCBs (Park et al., 2010).
In previous updates, chloracne and porphyria cutanea tarda were considered with the chronic non-cancer conditions. They are accepted as being associated with dioxin exposure, but they are considered acute outcomes and are no longer considered specifically in this chapter. For each type of health outcome, background information is followed by a brief summary of the findings described in earlier Veterans and Agent Orange (VAO) reports. In the discussion of the most recent scientific literature, the studies are grouped by exposure type (Vietnam veteran, occupational, or environmental). For articles that report on the health outcomes of a previously studied population, the detailed design information is summarized in Chapter 5. A synopsis of toxicologic and clinical information related to the biologic plausibility of the COIs influencing the occurrence of a health outcome is presented next and is followed by a synthesis of all the material reviewed. Each health-outcome section ends with the committee’s conclusions regarding the strength of the evidence that supports an association with the COIs. The categories of association and the committee’s approach to categorizing the health outcomes are discussed in Chapter 3.
For the purposes of this report, “non-cancerous respiratory disorders” are defined as all acute and chronic lung diseases other than cancers; the variety of conditions are described by the International Classification of Diseases (ICD), 9th Revision (ICD-9 460–519) or 10th Revision (ICD-10 J00–J99). Acute noncancerous respiratory disorders include pneumonia and other respiratory infections; they can increase in frequency and severity when the normal defense mechanisms of the lower respiratory tract are compromised.
Chronic non-cancerous respiratory disorders generally take two forms: airways diseases and parenchymal diseases. Airway diseases are disorders characterized by an obstruction of the flow of air out of the lungs; common examples
are asthma and chronic obstructive pulmonary disease (COPD). COPD includes such disorders as emphysema and chronic bronchitis. Parenchymal disease, or interstitial disease, generally includes disorders that cause inflammation and scarring of the deep lung tissue, including the air sacs and supporting structures. Parenchymal disease is less common than airway diseases and is characterized by a reduction in lung capacity, although it can also include a component of airway obstruction. Some severe chronic lung disorders, such as cystic fibrosis, are hereditary. Because Vietnam veterans received health screenings before entering military service, few severe hereditary chronic lung disorders are expected in that population.
More than 25 million people in the United States are thought to be living with asthma. As of 2015, the mortality rate for asthma among children and adults in the United States was highest among African Americans with 13.2 deaths per 100,000 in people less than 35 years old. Non-Hispanic whites had the second highest mortality rate, with 4.3 deaths per 100,000 in people less than 35 years old. There are nearly 14.8 million people who have physician-diagnosed COPD, but 12 million people are estimated to have undiagnosed COPD. The mortality rate for COPD was highest among non-Hispanic whites, with 127.8 deaths per 100,000 people. Native Americans had a death rate of 81.1 per 100,000 in individuals 45 years and older. African Americans ranked third, with a rate of 78.4 deaths per 100,000 people 45 years and older. Mortality rates were lowest among Asian or Pacific Islanders with 30.3 per 100,000 (healthypeople.gov, 2018b).
The most important risk factor for many non-cancerous respiratory disorders is the inhalation of cigarette smoke. Although exposure to cigarette smoke is not associated with every disease of the lungs, it is the major cause of many airway disorders, especially COPD; it also contributes to some interstitial disease, and it compromises host defenses in such a way that people who smoke are generally more susceptible to some types of pneumonia. Cigarette smoking also makes almost every respiratory disorder more severe and symptomatic than it would otherwise be. The incidence rates of habitual cigarette smoking vary with occupation, socioeconomic status, and generation. For those reasons, cigarette smoking can be a major confounding factor in interpreting the literature on risk factors for respiratory disease. Vietnam veterans are reported to smoke more heavily than non-Vietnam veterans (Kang et al., 2006; McKinney et al., 1997).
The causes of death from respiratory diseases, especially chronic diseases, are often misclassified on death certificates (Mieno et al. 2016). Grouping various respiratory diseases for analysis, unless they all are associated with a given exposure, will lead to an attenuation of the estimates of relative risk and to a diminution of statistical power. Moreover, the diagnosis of the primary cause of death from respiratory and cardiovascular diseases is often inconsistent. In particular, when a person had both conditions concurrently and both contributed to the death, there may be some uncertainty about which cause should be selected as the primary underlying cause. In other instances, errors may arise in selecting
one underlying cause in a complex chain of health events (for example, if COPD leads to congestive heart failure and then to respiratory failure).
Many study populations are small, so investigators often group deaths from all non-cancerous respiratory diseases into one category that combines pneumonia, influenza, and other diseases with COPD and asthma. The committee notes that an association between the group of all non-cancerous respiratory diseases with any of the COIs would be too nonspecific to be clinically meaningful; at most, such a pattern would be an indication that within this broad classification, the incidence of some particular disease entity might be affected by an exposure to a COI.
Conclusions from VAO and Previous Updates
The committee responsible for the initial VAO report (IOM, 1994) concluded that there was inadequate or insufficient information to determine whether there is an association between exposure to the COIs and respiratory disorders. Additional information available to the committees responsible for Update 1996 (IOM, 1996), Update 1998 (IOM, 1999), Update 2000 (IOM, 2001), Update 2002 (IOM, 2003c), and Update 2004 (IOM, 2005) also maintained the original conclusion of an inadequate or insufficient level of association.
A number of studies of non-malignant respiratory diseases in Vietnam veterans have since been reviewed. However, the majority of these studies were not able to control for major risk factors, such as smoking or tobacco use. Mortality from respiratory diseases was not found to be higher than expected in the Centers for Disease Control and Prevention’s Vietnam Experience Study (Boehmer et al., 2004), in the Air Force Health Study (AFHS) (Ketchum and Michalek, 2005), or in two Australian studies of Vietnam veterans (ADVA, 2005b,c), but it is possible that the use of death certificates may have introduced some misclassification of respiratory deaths (Drummond et al., 2010). In contrast, in the Army Chemical Corps (ACC) cohort of Vietnam veterans, the prevalence of self-reported noncancerous respiratory problems diagnosed by a doctor was significantly increased by about 40–60%, although no differences in the prevalence of respiratory problems were found in the subset of veterans whose serum TCDD was above 2.5 parts per trillion (Kang et al., 2006). Further study of cause-specific mortality in the ACC cohort by Cypel and Kang (2010) found a statistically significant excess of mortality from COPD when comparing the deployed and non-deployed groups. In a later analysis, Cypel and Kang were able to control for self-reported herbicide exposure, body mass index (BMI), and smoking status but found no statistically significant differences in respiratory diseases between sprayers and nonsprayers. Deaths due to COPD were lower in non-deployed ACC veterans than in males in the U.S. population; this is noteworthy because the prevalence of smoking in the non-deployed ACC veterans was about twice that in men in the U.S. population (Kang et al., 2006). Kang et al. (2014a) found lower risks
of respiratory system disease deaths and COPD deaths in women veterans who were deployed to Vietnam compared with non-deployed women veterans, but these associations were not statistically significant.
Comparing New Zealand veterans who served in Vietnam with the standardized general male population of New Zealand, the risk of death from all nonmalignant respiratory diseases (excluding COPD) was statistically significantly lower, but for COPD specifically no difference was found (McBride et al., 2013). Researchers of the Korean Veterans Health Study applied the Stellman exposure model and found no statistically significant difference for deaths from respiratory disease (Yi et al., 2014a,b). A comparison of deaths in the low-exposure category with the deaths in the high-exposure category found a statistically significant risk for non-malignant respiratory diseases that was statistically significantly elevated for COPD but not for pneumonia or asthma. A separate analysis of disease prevalence in the Korean Veterans Health Study used insurance claim data and found no difference in the risk of diseases of the respiratory system overall between the high-exposure and the low-exposure groups, but it did report statistically significant differences in the risks for COPD, pneumonia not due to influenza, chronic bronchitis, bronchiectasis, and asthma (Yi et al., 2014a).
Several studies of occupational cohorts have also been reviewed by the many VAO Update committees. Update 2000 (IOM, 2001) drew attention to findings in the Seveso cohort that suggested a higher mortality from non-cancerous respiratory disorders in study subjects, particularly males, who were more heavily exposed to TCDD. Additional follow-up of mortality in the Seveso cohort found some increase in mortality from COPD at the 15- and 20-year points (Bertazzi et al., 1998, 2001) as well as at the 25-year follow-up (Consonni et al., 2008).
Occupational and industrial cohorts in the United States, the United Kingdom, New Zealand, and Australia did not find increased mortality from non-cancerous respiratory diseases overall (Boers et al., 2010; Collins et al., 2009a,c; McBride et al., 2009a), though they were unable to control for smoking. An update of the National Institute for Occupational Safety and Health (NIOSH) cohort of pentochlorophenol (PCP) workers (Ruder and Yin, 2011) did not find an association between exposure to the COIs and deaths from all non-malignant respiratory diseases; no association with COPD deaths was found for the subgroup exposed to both TCDD and PCP. Updated mortality data on workers in two chlorphenoxy herbicide plants in the Netherlands were reanalyzed by Boers et al. (2012) using serum measurements, but no association of TCDD exposure with respiratory diseases was observed. An updated mortality study of workers in a pesticide factory with TCDD contamination (Manuwald et al., 2012) showed no association of non-malignant respiratory disease with exposure.
No associations were observed with respiratory mortality in a small subcohort of New Zealand phenoxy herbicide sprayers ('t Mannetje et al., 2005) or with mortality from COPD in private applicators or their spouses in the Agricultural Health Study (AHS) (Blair et al., 2005a). Several other studies using the
AHS cohort that examined morbidity from particular self-reported respiratory health problems, including wheeze, asthma, “farmer’s lung” or hypersensitivity pneumonitis, and chronic bronchitis were examined by update committees, but most findings were not statistically significant.
Based on the many studies considered in various cohorts that applied different designs, measures of exposure, and definitions of non-malignant respiratory outcomes, the conclusion of inadequate or insufficient evidence of an association between exposure to the COIs and mortality from all non-cancerous respiratory diseases or from COPD specifically, has remained unchanged. There also remains inadequate or insufficient evidence of an association between exposure to the COIs and the prevalence of respiratory diseases, such as wheeze or asthma, COPD, and farmer’s lung.
Update of the Epidemiologic Literature
This section summarizes the results of the relevant studies on respiratory disorders and exposures to the COIs. One new study of Vietnam veterans and respiratory disorders, one occupational study, and three other relevant studies of respiratory disorders and the COIs have been identified since Update 2014. Tables 43 and 44, which can be found at www.nap.edu/catalog/25137, summarize the results of studies related to non-cancer respiratory disease and COPD and pulmonary function, respectively.
Since Update 2014, one follow-up study of 2,783 male New Zealand Vietnam veterans, who served during 1964 to 1972 was identified and reviewed. Cox et al. (2015) used hospital discharge records from 1988 to 2009 to report the prevalent health conditions among this population. Age-specific hospitalization rates were calculated using the total number of annual hospitalizations published by the Ministry of Health and the average annual resident population. Standardized hospitalization rates and 99% confidence intervals (CIs) were calculated for both the veteran cohort and the general population, and a standardized hospitalization ratio (SHR) was calculated using the two rates. Results showed a statistically significant increase in the hospitalization risk for COPD (n = 300; SHR = 1.68, 99% CI 1.43–1.93) and pneumonia (n = 174; SHR = 1.36, 99% CI 1.09–1.63). No difference in the ratio was found for asthma, although the estimate was decreased (n = 33; SHR = 0.84, 99% CI 0.46–1.21). Overall, the results of this study showed a small increase in hospital admissions of Vietnam veterans for COPD and pneumonia compared with the standardized population of New Zealand. Because smoking is a major risk factor for respiratory conditions, the lack of smoking-adjusted ratios raises concerns about the validity of
the estimates. Moreover, exposure to the COIs was not validated and was simply assumed based on deployment to Vietnam.
Collins et al. (2016) added follow-up time for a retrospective cohort of 2,192 workers exposed to dioxins during trichlorophenol (TCP) and PCP production at chemical manufacturing plant in Michigan. Workers were compared with the U.S. population in order to calculate standardized mortality ratios, and work history records provided information about the length of the exposure. Serum samples used to measure levels of six types of dioxins were collected from 431 TCP and PCP workers. The historical concentrations for each dioxin congener were calculated based on the median concentration in the serum samples and the known half-lives associated with each congener. Complete vital status follow-up was achieved for the cohort, and there were 1,198 deaths during the entire study period (1979–2011). For the 1,198 TCP and PCP workers, there were 110 deaths from respiratory system disorders, with an SMR of 0.94 (95% CI 0.77–1.13).
Henneberger et al. (2014) examined the exacerbation of asthma among pesticide applicators with asthma who were enrolled in the AHS; details of the AHS study design, data collection, and cohort are found in Chapter 5. In this analysis, participants were selected for inclusion based on completing both the baseline and an additional take-home questionnaire and having reported a doctor diagnosis of asthma and also having reported active asthma based on having had at least one episode of wheezing or whistling in the previous 12 months and having had breathing problems in the same time period. The final study sample included 926 adult pesticide applicators with active asthma. Exacerbation was defined as having visited a hospital emergency room or doctor for an episode of wheezing or whistling in the previous 12 months. AHS participants were asked about their use of individual pesticides, and this analysis focused on the 36 pesticides used in the 12 months before enrollment (current exposure) by at least 10 applicators with at least two exacerbation cases. Logistic regression was used to estimate odds ratios for pesticide exposure, controlling for age, state, type of pesticide applicator (private or commercial), cigarette smoking status, allergy status based on self-reports of doctor-diagnosed hay fever or eczema, and adult onset of asthma based on onset at >20 years of age. No association with an elevated odds of asthma exacerbation was found for either 2,4-D (OR = 0.8; 95% CI 0.5–1.3) or dicamba (OR = 1.0; 95% CI 0.6–1.6). Interaction models for pesticide exposure and allergic status were not significant for 2,4-D and dicamba. The authors interpreted the inverse associations as related to the possibility that asthmatic farmers avoided certain activities and exposures. The AHS includes a well-characterized cohort with self-reported pesticide exposure data that has been shown to be reliable and the asthma and exacerbation outcomes were carefully
defined. However, the data are cross-sectional, and the sequence of exposure and events cannot be determined.
't Mannetje et al. (2018) conducted a morbidity survey among a subset of workers who were employed at the New Plymouth, New Zealand, phenoxy herbicide production plant for at least 1 month between 1969 and 1984. The plant produced 2,4,5-T, and workers were potentially exposed to 2,4,5-T, the intermediates of TCP and other chlorophenols, and TCDD. Workers had previously been recruited and examined as part of the international cohort of producers of phenoxy herbicides led by the International Agency for Research on Cancer (IARC) (Kogevinas et al., 1997); see Chapter 5 for more details on the IARC cohort and the New Zealand phenoxy producers. This study extends the follow-up period of these workers to approximately 30 years from their last 2,4,5-T production exposure. From the original cohort of 1,025 workers, 631 were living, had a current address in New Zealand, and were below 80 years of age on January 1, 2006. For the current follow-up, 430 of the 631 workers were randomly selected and invited to participate in the morbidity survey, of which 245 (57%) participated. The survey was administered in 2007–2008 by face-to-face interview, and information was collected on demographic factors and health information, including doctor-diagnosed conditions and the year of diagnosis. A blood sample was also collected at that time and analyzed for TCDD, lipids, thyroid hormones, and other substances. For 111 participants, a neurological examination was conducted. Associations between exposure and health outcomes were assessed using logistic regression models that controlled for age, gender, smoking, BMI, and ethnicity using two different methods of exposure: having worked in a TCDD-exposed job (based on occupational records) and having serum TCDD concentration ≥10 pg/g lipid (18%). Mean TCDD concentrations were 19 pg/g lipid in the 60 men directly involved in phenoxy/TCP production, and 6 pg/g lipid in the 141 men and 43 women who worked in other parts of the plant. Compared with the 124 people in the non-highly-exposed jobs, the 121 people who had ever worked in a highly exposed job were no more likely to have doctor-diagnosed asthma (n = 8; OR = 1.13; 95% CI 0.40–3.22) or chronic bronchitis (n = 3; OR = 2.39; 95% CI 0.17–34.0). When compared by serum TCDD concentration ≥10 pg/g lipid, there were few cases of asthma or chronic bronchitis, and no difference in risk was found. Diagnoses of tuberculosis, pleurisy, or pneumonia were also examined, but there were too few cases to present valid estimates.
Cappelletti et al. (2016) performed a retrospective study of 331 male electric arc foundry workers at a single plant in Trentino, Italy, to determine if they experienced excess mortality from all causes or were at an increased risk for several other diseases, including asthma, due to occupational exposures to foundry dust. Analysis of the dust emissions found that the dust contained metals (including iron, aluminum, zinc, manganese, lead, chromium, nickel, cadmium, mercury, and arsenic), polycyclic aromatic hydrocarbons (PAHs), PCBs, and
polychlorinated dibenzo-p-dioxin/dibenzofurans (PCDD/Fs) (reported as TEQs). Therefore, the authors could not determine which of the agents were associated with a specific outcome or to what extent. The men had worked at the factory for at least 1 year and, for the asthma analysis, were compared with 32 presumed non-exposed workers (clerks, managers, and watchmen) or with the standardized general population of Region Trentino-Alto Adige (where the factory was located) because there were few non-exposed foundry workers and high attrition rates. Company and medical records were used to determine vital status; the cause of death was determined from death certificates or other registries. Requests for exemption health care fees were used as a surrogate measure to identify the most common morbid conditions in the general population, which were then applied to the cohort to compute the relative risks for each of the conditions. The workers were followed from March 19, 1979 (or their first day of employment), through December 31, 2009, or date of death. The analysis for asthma was limited to 235 workers, and effect estimates were calculated using Mantel-Haenszel tests. Only two cases of asthma were found among the workers, and there was no difference in risk compared with the age-adjusted provincial population, although the effect estimate was imprecise (RR = 1.08, 95% CI = 0.27–4.31). This study is most limited by the fact that foundry dust is a complex mixture, which made it impossible to discern the impact of the specific contaminants of the foundry dust on the health outcomes of those exposed workers. Estimates were adjusted for only the age group and were not adjusted for other risk factors such as tobacco use, BMI, or other jobs or activities that could result in similar exposures. Exposure to foundry dust by the general population that was used for comparison was not discussed, although the foundry appears to be in the local vicinity, and emissions from it were reported to be present within a 2-kilometer radius.
Other Identified Studies
Five other studies of non-malignant respiratory diseases were identified, but all lacked sufficient exposure specificity to be included as contributing to the evidence base of the potential effect of the COIs. One study of U.S. capacitor manufacturer workers exposed to mixed PCBs that reported on deaths from respiratory diseases was identified, but it was limited by its lack of exposure specificity (Ruder et al., 2014). An analysis of data from the AGRICAN French cohort study examined the prevalence of respiratory diseases in adult men and women who were active or retired farm workers or owners, but all exposures and outcomes, which included asthma, were self-reported, and exposures to specific pesticides were not ascertained (Baldi et al., 2014). Butinof et al. (2015) conducted a cross-sectional study of male pesticide applicators who were licensed in Argentina and who were directly exposed to pesticides (n = 880). All participants completed a self-administered questionnaire that was adapted from
the U.S. AHS, which collected information about herbicides, insecticides, and fungicides used, but no pesticide-specific exposure assessment was conducted, which limited the study’s utility. A study of occupational exposure to pesticides in Ethiopian farmers and farm workers and respiratory health effects was conducted by Negatu et al. (2016) but was not considered further since data on specific pesticides were not collected. Finally, Akahane et al. (2017) examined the prevalence of self-reported long-term health effects (including respiratory disorders) in people exposed to PCBs, dioxins (e.g., PCDD/Fs), and dioxin-like chemicals through the ingestion of contaminated rice bran oil (Yusho accident) compared with an age-, sex- and residential-area-matched group. Because no TEQs or other quantification of relevant exposures were presented, the study was not considered further.
Several recent studies explored the pro- and anti-inflammatory activity of TCDD exposure and of aryl hydrocarbon receptor (Ahr) in lungs in mice and in mouse and human lung cells. These studies addressed such health effects as lung response to viral infections, asthma, COPD, and allergen-induced lung inflammation. Cho et al. (2015) found that nuclear coactivator 7 (NCOA7) showed differential activity in a cell-culture study of TCDD exposure in normal lung epithelial and lung cancer cells with NCOA7 increasing CYP1A1 activity and the levels of inflammatory cytokines in the diseased cells but not in normal cells. S. M. Lee et al. (2017) found that AHR activation by the endogenous ligand kynurenine plays a role in immune regulation in inflammatory responses in a mouse model of lung injury.
Q. Liu et al. (2014) exposed juvenile zebrafish to TCDD in diet and reported a novel finding of lesions in nasal epithelium. Boule et al. (2015) exposed mice to TCDD during development and investigated their response to viral infection as adults. Treated mice had increased bronchopulmonary inflammation with an enhanced CD4+ T-cell response.
X. M. Li et al. (2016) found that TCDD-Ahr activation prevented inflammation and airway hyper-responsiveness in a non-eosinophilic asthma model in mice. The lungs of treated mice showed reduced eosinophil and neutrophil infiltration, and Th17 cytokine IL-17 was reduced in serum and bronchio-alveolar lavage fluid (with Th2 cytokine IL-4 also reduced). In a study comparing Ahr knockout and wild-type mice, T. Xu et al. (2015) found that TCDD treatment enhanced mesenchymal stem cell recruitment, thereby suppressing allergen-induced lung inflammation.
Sarill et al. (2015) explored the role of AHR and suggested that AHR has a role regulating anti-oxidant defenses in lung structural cells. In Ahr knockout mouse lung fibroblasts and human lung adenoma cells that were deficient in AHR expression and exposed to cigarette smoke extract, they found increased
oxidative stress compared to wild-type cells. They also showed that anti-oxidant gene induction was significantly less in the Ahr knockout mouse lung fibroblasts. In lung fibroblasts from mice with an Ahr mutation (incapable of DNA binding), they showed that anti-oxidant sulfiredoxin 1 induction after cigarette smoke-extract exposure was independent of dioxin-response element binding (i.e., a role for AHR independent of exogenous ligand). In sum, the findings of Sarill et al. (2015) suggest that low AHR expression may facilitate the development or progression of COPD. Zago et al. (2017) studied the role of AHR and Cox-2 protein suppression and also concluded that low AHR contributes to increased inflammation, based on their study in human lung fibroblasts.
Update 2014 included new data that did not add any compelling evidence of an association of exposure to the COIs and respiratory disorders. The current update still does not display a coherent body of epidemiologic evidence to support conclusions concerning whether the risks of other particular respiratory problems are associated with exposure to the COIs. Cox et al. (2015) showed an increase in hospitalization risk rates for COPD and pneumonia, but not asthma, in New Zealand Vietnam veterans. However, smoking was not controlled for in the analysis and exposure to the COIs was not verified. In a 30-year follow-up of workers who produced 2,4,5-T in New Zealand, highly exposed workers did not have an elevated risk of doctor-diagnosed asthma, chronic bronchitis, tuberculosis, pleurisy, or pneumonia compared with lower-exposed workers based on job duties or serum TCDD concentration. The study of electric arc foundry workers in Italy (Cappelletti et al., 2016) reported no difference in the risk of asthma, but this was based on two cases. Moreover, the primary exposure, foundry dust, is a complex mixture, and it is not possible to isolate any potential COI exposure. In a follow-up of TCP and PCP workers at the Dow Midland, Michigan, plant, Collins et al. (2016) did not find elevated mortality for respiratory system disorders. Henneburger et al. (2014) found no association of asthma exacerbation with exposure to the COIs in the AHS. Therefore, the prior assessment cannot be altered since the new findings are mixed, and the study designs have limitations.
On the basis of the evidence reviewed here and in previous VAO reports, the committee concludes that there is inadequate or insufficient evidence of an association between exposure to the COIs and mortality from all non-cancerous respiratory diseases or from COPD specifically. There is also inadequate or insufficient evidence of an association between exposure to the COIs and the prevalence of respiratory disorders.
This section discusses a variety of conditions specified by ICD-9 520–579 or ICD-10 K00–K95: diseases of the esophagus, stomach, intestines, rectum, liver, and pancreas. Details on peptic ulcer and liver disease, the two conditions most often discussed in the literature reviewed, are provided below. The symptoms and signs of gastrointestinal disease and liver toxicity are highly varied and often vague.
The essential functions of the gastrointestinal tract are to absorb nutrients and to eliminate waste. Those complex tasks involve numerous chemical and molecular interactions on the mucosal surface and complex local and distant neural and endocrine activity. One common condition of the gastrointestinal tract is motility disorder, which is present in about 15% of adults. The most convenient way to categorize diseases that affect the gastrointestinal system is according to the affected anatomic segment. Esophageal disorders predominantly affect swallowing, gastric disorders are related to acid secretion, and conditions that affect the small and large intestines are reflected in alterations in nutrition, mucosal integrity, and motility. Some systemic disorders (inflammatory, vascular, infectious, and neoplastic conditions) also affect the gastrointestinal system.
Peptic-ulcer disease refers to ulcerative disorders of the gastrointestinal tract that are caused by the action of acid and pepsin on the stomach or duodenal mucosa. Peptic-ulcer disease is characterized as gastric or duodenal ulcer, depending on the site of origin. Peptic-ulcer disease occurs when the corrosive action of gastric acid and pepsin overcomes the normal mucosal defense mechanisms that protect against ulceration. About 10% of the population has clinical evidence of having had a duodenal ulcer at some time in their lives; a similar percentage is affected by gastric ulcer. The incidence of duodenal ulcer peaks in the fifth decade, and the incidence of gastric ulcer about 10 years later.
Evidence increasingly indicates that the bacterium Helicobacter pylori is linked to peptic-ulcer disease (both duodenal and gastric). H. pylori colonizes the gastric mucosa in 95–100% of patients who have a duodenal ulcer and in 75–80% of patients who have a gastric ulcer. Healthy people in the United States under 30 years old have gastric colonization rates of about 10%. Over the age of 60 years, colonization rates exceed 60%. Colonization alone, however, is not sufficient for the development of ulcer disease; only 15–20% of subjects who have H. pylori colonization will develop ulcers in their lifetime. Other risk factors include genetic predisposition (such as some blood and human leukocyte antigen types), cigarette smoking, and psychologic factors (chronic anxiety and stress).
Blood tests of liver function are the mainstay of diagnosing liver disease. Increases in serum bilirubin and in the serum concentrations of some hepatic enzymes—aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, and γ-glutamyltransferase (GGT)—are commonly noted in liver disorders. The relative sensitivity and specificity of those enzymes for diagnosing liver disease vary, and a diagnosis can require several tests. The only regularly reported abnormality in liver function associated with TCDD exposure in humans is an increase in GGT. The estimated serum activity of that enzyme is a sensitive indicator of a variety of conditions, including alcohol and drug hepatotoxicity, infiltrative lesions of the liver, parenchymal liver disease, and biliary tract obstruction. Increases are noted after many chemical and drug exposures that are not followed by evidence of liver injury. The confounding effects of alcohol use (often associated with increased GGT) make it difficult to interpret changes in GGT in exposed people (Calvert et al., 1992). An increase in GGT can be considered a normal biologic adaptation to chemical, drug, or hormone exposure.
Cirrhosis is the most commonly reported liver disease in epidemiologic studies of herbicide or TCDD exposure. Cirrhosis is an irreversible chronic injury of the liver with extensive scarring and a resulting loss of function. Clinical symptoms and signs include jaundice, edema, abnormalities in blood clotting, and metabolic disturbances. Cirrhosis can lead to portal hypertension with associated gastroesophageal varices, an enlarged spleen, abdominal swelling attributable to ascites, and, ultimately, hepatic encephalopathy that can progress to coma. It is generally impossible to distinguish the various causes of cirrhosis through the clinical signs and symptoms or pathologic characteristics. The most common cause of cirrhosis in North America and many parts of western Europe and South America is excessive alcohol consumption. Other causes are chronic viral infection (hepatitis B or hepatitis C), the poorly understood condition primary biliary cirrhosis, chronic right-sided heart failure, and a variety of less common metabolic and drug-related conditions.
Conclusions from VAO and Previous Updates
Several studies reviewed by previous VAO committees have reported on several non-cancerous digestive system outcomes, and for that reason this section and the update of new epidemiologic literature presents all results for digestive and hepatobiliary outcomes by publication. Some studies that have been reviewed focused on liver enzymes, while others reported on specific liver diseases. An evaluation of the effects of herbicide and TCDD exposure on non-cancerous gastrointestinal ailments is challenging in that clinical experience suggests that medical history and physical examination are undependable diagnostic tools for some ailments, so incidence data are sometimes problematic. The strong
interdependence among the characteristics of a given person (such as weight and the laboratory indexes of hepatic function and health) and TCDD body burden complicates the already difficult task of assessing association.
Studies of Vietnam veterans have generally examined different laboratory endpoints and clinical conditions, making comparisons and overall conclusions difficult. An analysis of the AFHS found that a significantly higher percentage of Ranch Hand veterans in the high-dioxin category had excesses of transaminase and other nonspecific laboratory measures of liver function than their comparison subjects. The data were consistent with an interpretation of a dose–response relationship, but other explanations were also plausible (AFHS, 2000). Later analyses also reported some abnormalities in liver enzymes among the Ranch Hands, including decreasing levels of C4 complement as dioxin increased and abnormal triglyceride concentrations that increased as the 1987 dioxin concentration increased (AFHS, 2005). However, mortality studies of the Ranch Hand cohort have not found an increased mortality related to gastrointestinal or liver disease (Ketchum and Michalek, 2005). Among ACC veterans, an increased rate of hepatitis was associated with Vietnam service but not with a history of spraying herbicide (Kang et al., 2006), and an 80% excess of digestive system or cirrhosis deaths was observed in veterans who handled or sprayed herbicides in Vietnam compared with their non-Vietnam-veteran peers (Cypel and Kang, 2010).
Studies of digestive and liver disease among other Vietnam veterans cohorts have not found associations with presumed herbicide exposure. The Australian Vietnam-veterans study (ADVA, 2005b) did not find an increase in liver disease in military personnel who served in Vietnam compared with the general population of Australia. The relationship between possible herbicide exposure and liver and gastric-ulcer disease was described in a sample of Korean Vietnam-era veterans in three publications by Yi et al. (2013, 2014a,b). Based on health insurance claims data, the adjusted prevalence of peptic-ulcer disease was found to be 3% higher in those with high exposure than in those with low exposure after adjusting for several behavioral, demographic, and service-related factors. For liver cirrhosis, there was likewise a small elevation in the prevalence and a significant log-linear relationship between an exposure opportunity score and the odds of having cirrhosis. In the mortality study in the same cohort of Korean Vietnam veterans, there was no association between putative log-transformed exposure and mortality from peptic ulcers (Yi et al., 2014b). However, highly exposed veterans were 17% more likely to die from cirrhosis than those with low exposure. Deaths from alcoholic liver disease were also statistically significantly elevated in the more highly exposed veterans.
Most of the analyses of occupational or environmental cohorts have had insufficient numbers of cases to support confident conclusions. Gastric ulcer was not associated with serum levels of dioxin-like PCDD/Fs or dioxin-like PCBs or with total TEQs after adjusting for possible confounders in a general population sample of 2,264 Japanese adults (Nakamoto et al., 2013). An analysis of National
Health and Nutrition Examination Survey (NHANES) data on blood levels of 37 environmental pollutants and alanine aminotransferase levels in 1,345 persons aged 12 years and older found that, in general, blood levels of dioxin-like chemicals were not significantly correlated with alanine aminotransferase levels. The exception was 1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin, which was associated with only a slight elevation (< 1%) (Yorita Christensen et al., 2013).
A study of the IARC cohort of phenoxy-herbicide and chlorophenol production workers and sprayers (Vena et al., 1998), the only study that had a relatively large number of observations, found less digestive system disease and cirrhosis mortality in exposed workers than in non-exposed controls. The mortality results through 2001 for the Seveso, Italy, cohort (Consonni et al., 2008) found no excess of deaths related to digestive diseases or specifically to cirrhosis. Several mortality studies of various occupational cohorts exposed to the COIs have been inconsistent but generally found no statistically significant increases in deaths from either ulcers or cirrhosis (Boers et al., 2010, 2012; Collins et al., 2010a,b; Manuwald et al., 2012; McBride et al., 2009a,b; Ruder and Yiin, 2011).
Thus, the reports have been inconsistent, and interpreting individual studies is difficult because of a lack of information on alcohol consumption and other risk factors. In the studies that showed the strongest association between potential exposure and gastrointestinal disease (specifically cirrhosis), there was strong evidence that excess alcohol consumption was the cause of the cirrhosis. The committee responsible for VAO concluded that there was inadequate or insufficient information to determine whether there is an association between exposure to the COIs and gastrointestinal and digestive disease, including liver toxicity. Additional information available to the committees responsible for subsequent updates has not changed that conclusion.
Update of the Epidemiologic Literature
One new study of gastrointestinal diseases among Vietnam veterans and two occupational studies of workers in herbicide-producing plants have been identified since Update 2014.
Cox et al. (2015) used hospital discharge records from 1988 to 2009 to identify health conditions that affected 2,783 male New Zealand veterans who had served in Vietnam during 1964–1972. Age-specific hospitalization rates were calculated using the total number of annual hospitalizations published by the Ministry of Health and the average annual resident population. Standardized hospitalization rates and 99% CIs were calculated for the veteran cohort and the general population and reported for 14 conditions related to gastrointestinal and hepatobiliary outcomes. Those conditions were: esophageal disorders,
gastroduodenal ulcer, gastritis, appendicitis, abdominal hernia, intestinal obstruction, diverticulosis, anal/rectal disorders, biliary disorders, alcoholic liver disease, other liver disease, pancreatic disease, abdominal pain, and gastrointestinal hemorrhage. The risk of hospitalization for gastroduodenal ulcer was elevated for veterans (n = 44; SHR = 1.80, 99% CI 1.10–2.50). Few cases of alcoholic liver disease were found, but the risk of hospitalization was not different between veterans and the general population (n = 11; SHR = 1.07, 99% CI 0.24–1.90). The risk of hospitalization from other liver disease (not further defined) was higher in veterans (n = 48; SHR = 1.86, 99% CI 1.17–2.55). No other increased or decreased risks of gastrointestinal and hepatobiliary outcomes were found. Exposure to the COIs was not validated and was assumed based on deployment to Vietnam. Moreover, the analysis did not control for smoking or ethnicity or other potentially important risk factors.
Collins et al. (2016) included additional follow-up time in a retrospective analysis of a cohort of 2,192 workers exposed to dioxins during TCP and PCP production at chemical manufacturing plant in Michigan. The U.S. population was used as the comparator for estimating standardized mortality ratios. Work history records were used to determine the length of the exposure. Serum samples for measuring the levels of six types of dioxins were collected for 431 TCP and PCP workers. The historical concentrations for each dioxin congener were calculated from the median concentrations from the serum samples and the known half-lives associated with each congener. Complete vital status follow-up was achieved for the cohort, and there were 1,198 deaths through the entire study period (1979–2011). Deaths from ulcer of the stomach and duodenum (K25–K27) and cirrhosis of the liver (K70–K74) were also included. No difference in the mortality of ulcer of the stomach and duodenum was found for the TCP workers (n = 2; SMR = 0.78, 95% CI 0.10–2.82), but an increased risk was found for the PCP workers (n = 5; SMR = 3.38, 95% CI 1.10–7.89). However, the interpretation of this finding is limited because the number of deaths from ulcers was small and the resulting risk estimates were imprecise. There were 16 total reported deaths from liver cirrhosis, but no difference in mortality was found for the TCP workers (n = 8; SMR = 0.44, 95% CI 0.19–87.2) or the PCP workers (n = 8; SMR = 0.97, 95% CI 0.42–1.91) compared with the standardized U.S. population. The committee notes that there is a possible mistake in the confidence interval reported for the TCP risk estimate, but there is no compelling data to indicate that cirrhosis of the liver is associated with exposure to dioxin. The committee also notes that no new deaths from stomach or duodenum ulcer have been reported since the last update of this cohort in 2003 (Collins et al., 2009a,b) at which time a statistically significant increase in stomach and duodenal ulcer deaths was found
for PCP workers but not TCP workers. Since 2003, two additional deaths from liver cirrhosis among TCP workers were reported.
't Mannetje et al. (2018) conducted a morbidity survey among a subset of workers who were employed at the New Plymouth, New Zealand, phenoxy herbicide production plant for at least 1 month between 1969 and 1984. The plant produced 2,4,5-T, and workers were potentially exposed to 2,4,5-T, the intermediates of TCP and other chlorophenols, and TCDD. Workers had previously been recruited and examined as part of the international cohort of producers of phenoxy herbicides led by IARC (Kogevinas et al., 1997); see Chapter 5 for more details on the IARC cohort and the New Zealand phenoxy producers. This study extends the follow-up period of these workers to approximately 30 years from their last 2,4,5-T production exposure. From the original cohort of 1,025 workers, 631 were living, had a current address in New Zealand, and were below 80 years of age on January 1, 2006. For the current follow-up, 430 of the 631 workers were randomly selected and invited to participate in the morbidity survey, of which 245 (57%) participated. The survey was administered in 2007–2008 by face-to-face interview, and information was collected on demographic factors and health information, including doctor-diagnosed conditions and the year of diagnosis. A blood sample was also collected at that time and analyzed for TCDD, lipids, thyroid hormones, and other substances. For 111 participants, a neurological examination was conducted. Associations between exposure and health outcomes were assessed using logistic regression models that controlled for age, gender, smoking, BMI, and ethnicity using two different methods of exposure: having worked in a TCDD-exposed job (based on occupational records) and having serum TCDD concentration ≥10 pg/g lipid (18%). Mean TCDD concentrations were 19 pg/g lipid in the 60 men directly involved in phenoxy/TCP production and 6 pg/g lipid in the 141 men and 43 women who worked in other parts of the plant. Compared with the 124 people in the non-highly-exposed jobs, the 121 people who had ever worked in a highly exposed job were no more likely to have a doctor-diagnosed liver function problem (n = 10; OR = 0.84; 95% CI 0.30–2.34). When compared by serum TCDD concentration ≥10 pg/g lipid, there were few cases of liver function problem, leading to imprecise effect estimates (n = 3; OR = 0.54; 95% CI 0.12–2.36).
Other Identified Studies
Ruder et al. (2014) assessed cause-specific mortality including grouped diseases of the digestive system among 24,865 workers exposed to dioxin-like and non-dioxin-like PCBs from three U.S. electrical capacitor manufacturing plants who were employed 3 months or more from as early as 1939 through 1977. However, a lack of exposure specificity precluded further consideration. Akahane et al. (2017) examined the prevalence of self-reported long-term health effects in people exposed to PCBs, dioxins (e.g., PCDD/Fs), and dioxin-like chemicals through the ingestion
of contaminated rice bran oil (Yusho accident) compared with an age-, sex- and residential-area-matched group. Because no TEQs or other quantification of the relevant exposures was presented, the study was not considered further.
Yamamoto et al. (2015) studied 678 male workers at a waste incineration plant in Japan. Blood samples were measured to determine the amounts of TCDD and PCBs in the blood, and several markers of hematology and blood chemistry profiles were also measured. Although some of these tests may be able to detect liver dysfunction, it was not linked to a health outcome, and the study was not considered further.
The liver is the first organ that encounters chemicals absorbed from the gastrointestinal tract, and it is responsible for metabolizing them to water-soluble chemicals that can be excreted in the urine. Because the liver has many detoxifying enzymes that efficiently metabolize many chemicals, liver toxicity is usually associated only with high-dose acute exposure or lower-dose chronic exposure. The liver can be damaged if the metabolism of a chemical results in the production of a reactive intermediate that is more toxic than the parent chemical. Changes in the serum concentrations of liver enzymes are biomarkers of liver toxicity, and their magnitudes correlate with the degree of liver damage. The exposure of laboratory animals to high doses of 2,4-D, 2,4,5-T, and TCDD is known to cause liver damage. The mechanisms by which the phenoxy herbicides damage the liver are based on the inhibition of mitochondrial function by the blocking of oxidative phosphorylation; this leads to a loss of generation of adenosine triphosphate, the death of cells, and hepatic necrosis and fibrosis. TCDD-induced hepatotoxicity is mediated by the activation of the AHR, which leads to changes in gene transcription and associated changes in cell function. Changes in gene expression are associated with several physiologic processes, oxidative stress, and apoptosis (Boverhof et al., 2005, 2006). TCDD-mediated hepatic steatosis is characterized by the accumulation of triglycerides caused by the combined up-regulation of CD36/fatty acid translocase and fatty acid transport proteins, the suppression of fatty acid oxidation, the inhibition of the hepatic export of triglycerides, an increase in peripheral fat mobilization, and an increase in hepatic oxidative stress (J. H. Lee et al., 2010). Recent evidence suggests that hepatic steatosis produced by TCDD might be mediated by the mitochondria (He et al., 2013). The exposure of rats to TCDD over a 2-year period (NTP, 2004) also produced several changes in the liver, including hepatocyte hypertrophy, multinucleated hepatocytes, inflammation, pigmentation, diffuse fatty change, necrosis, bile duct hyperplasia, bile duct cyst, nodular hyperplasia, portal fibrosis, and cholangiofibrosis. Lamb et al. (2016) studied a mouse model of liver injury and found that TCDD exposure increased
liver damage and inflammation, suggesting that TCDD-activated AHR enhanced hepatic stellate cell activity.
AHR displays species differences; for example, amino acid sequences in the C-terminal region of human and mouse AHR are only 58% identical. Compared with the mouse Ahr, the human AHR has about a 10-fold lower relative affinity for TCDD; the difference has been attributed to the amino acid residue valine 381 in the ligand-binding domain of the human AHR (Flaveny et al., 2009; Ramadoss and Perdew, 2004). The existence of species differences associated with AHR activation is supported by the divergence in the transcriptomic and metabolomic responses to TCDD in mouse, rat, and human liver (Boutros et al., 2008, 2009; Carlson et al., 2009; Forgacs et al., 2012, 2013; Kim et al., 2009; Nault et al., 2013; Watson et al., 2017). In a recent study, gene-expression changes were compared in adult female primary human and rat hepatocytes that had been exposed to TCDD in vitro (Black et al., 2012). Whole-genome microarrays found that TCDD produced divergent gene-expression profiles in rat and human hepatocytes, both on an ortholog basis (conserved genes in different species) and on a pathway basis. For commonly affected orthologs or signaling pathways, the human hepatocytes were about 15-fold less sensitive than rat hepatocytes. Another microarray study examining species-specific transcriptomic differences in primary hepatocytes from humans, mice, and rats identified 16 orthologous genes that were dysregulated by TCDD in all three species (Forgacs et al., 2013). Such findings are consistent with epidemiologic studies that have shown humans to be less sensitive to TCDD-induced hepatotoxicity. However, it should be noted that in vitro human hepatocyte studies may not reflect the in vivo response of human liver to TCDD.
There have been few reports of health-relevant effects of phenoxy herbicides or TCDD on the gastrointestinal tract, even after high exposure. A series of recent papers demonstrated the influence of TCDD and the Ahr pathway on the development of the gut microbiome in mice (Stedtfeld et al., 2017a,b,c), which is thought to influence immune response that may affect gastrointestinal disorders. However, the available animal data do not support a plausible link between herbicide exposure and gastrointestinal toxicity in Vietnam veterans.
Previous updates indicated that there have been inconsistent findings for the COIs and gastrointestinal and liver disease and that interpreting individual studies is difficult because of a lack of information on alcohol consumption (especially for liver cirrhosis) and other risk factors. The recent studies of Vietnam veterans from New Zealand (Cox et al., 2015) and occupational cohorts of phenoxy herbicide producers (Collins et al., 2016; 't Mannetje et al., 2018) have not provided informative evidence for an association with the COIs. Although a few animal studies have shown an influence of TCDD and the Ahr pathway on the development of the gut microbiome in mice (Stedtfeld et al., 2017a,b,c), which may
modify the immune response in gastrointestinal disorders. The available animal data do not support a plausible link between herbicide exposure and gastrointestinal toxicity in Vietnam veterans.
On the basis of the evidence reviewed here and in previous VAO reports, the committee concludes that there is inadequate or insufficient evidence to determine whether there is an association between exposure to the COIs and gastrointestinal and digestive diseases.
Update 2014 was the first update for which the literature search identified studies reporting results concerning a possible association between exposure to the COIs and kidney diseases (ICD-9 580–589; ICD-10 N00–N29). The kidneys are located in the lower back region; their main function is to filter wastes and excess water out of the blood, which results in the production of urine. The kidneys are also responsible for helping maintain the body’s chemical balance, helping control blood pressure, and synthesizing hormones. When problems arise with kidney function, it is often the result of damaged nephrons, which may leave the kidneys unable to filter blood and, thus, unable to remove wastes, which can then accumulate in the body. Chronic kidney disease is characterized by a gradual and usually permanent loss of kidney function that often results in renal failure. Diabetes, hypertension, and glomerulonephritis (acute inflammation) can all increase the risk of kidney disease.
Conclusions from VAO and Previous Updates
Publications from the Korean Veterans Health Study included findings for non-malignant kidney disease. Yi et al. (2013a) examined the prevalence of self-reported exposure based on six questions and self-reported kidney failure using a postal survey of 114,562 Korean veterans who had served in the Vietnam War. The incidence of kidney failure was compared by defining high and low categories of exposure based on the six exposure questions and also by exposure opportunity index (EOI) scores that were calculated. Self-reported kidney failure was statistically significantly increased when the analysis was based on perceived exposures, but it was not significant when the analysis was based on the EOI scores. In a study of cause-specific mortality through 2005 for 180,639 Korean veterans who were alive in 1992, Yi et al. (2014b) found that after adjustment for age in 1992 and rank, no differences in the hazard ratios were observed for acute renal failure or for chronic renal failure when the analyses were based on the EOI scores.
Two environmental studies of non-U.S. populations were also considered. In a cross-sectional study of 2,264 Japanese men and women who had not been occupationally exposed to dioxins, self-reported kidney disease (not otherwise specified) was not associated with serum levels of dioxin-like PCDD/Fs or dioxin-like PCBs or with total TEQs after adjusting for possible confounders (Nakamoto et al., 2013). The second study sought to determine the factors contributing to a form of kidney disease not related to diabetes, hypertension, or any other recognized cause in adults in Sri Lanka (Jayatilake et al., 2013). Urine samples were taken from 57 cases and from 39 controls who were from non-endemic areas, and the samples were analyzed for 11 biomarkers of pesticides, including the COIs 2,4-D, 2,4,5-T, and 2,4,5-TCP. Of these, only 2,4-D was among the seven biomarkers found at concentrations above the limit of detection; 3.5% of the cases had 2,4-D concentrations above the reference limit of 0.3 μg/L. Since the urinary pesticide results were presented for only the cases, no inference can be made about the relative risk for this kidney condition in association with 2,4-D.
Based on these four studies, the committee for Update 2014 concluded that there was inadequate or insufficient evidence of exposure to the COIs and nonmalignant kidney diseases.
Update of the Epidemiologic Literature
One new study of Vietnam veterans and kidney and urinary disorders has been identified since Update 2014, as well as supporting occupational, environmental, and case-control studies of kidney and urinary disorders and exposure to the COIs.
Cox et al. (2015) used hospital discharge records from 1988 to 2009 to identify health conditions that affected 2,783 male New Zealand veterans who had served in Vietnam during 1964–1972. Age-specific hospitalization rates were calculated using the total number of annual hospitalizations published by the Ministry of Health and the average annual resident population. Standardized hospitalization rates and 99% CIs were calculated for the veteran cohort and the general population and the two rates were used to calculate an SHR for several conditions including kidney and urinary outcomes. An elevation in chronic renal failure risk was identified (SHR = 1.21, 95% CI 1.07–1.36), with an acceleration in the risk for later time periods. No significant associations were identified with urinary tract infections (OR = 1.06, 95% CI 0.66–1.46), benign prostatic hypertrophy (OR = 1.11, 95% CI 0.79–1.42), or urinary stones (OR = 1.06, 95% CI 0.81–1.31). The exposures were not validated through serum measurements and
were assumed based on deployment to Vietnam, and the study did not control for smoking status, ethnicity, or other potentially important risk factors.
Two studies using data from the AHS were reviewed. Using a prospective cohort design with an average of 16 years of follow-up, Lebov et al. (2016) evaluated the association between use of 41 specific pesticides and end-stage renal disease in 55,580 male pesticide applicators. Significant associations were found with several non-COIs (p for trend < 0.05), but there were no statistically significant associations with the phenoxy herbicides (2,4-D p for trend = 0.32; 2,4,5,T p for trend = 0.55). An education level greater than high school and obesity at enrollment were also associated with end-stage renal disease, as were diabetes, high blood pressure, and kidney disease.
Lebov et al. (2015) evaluated the use of 50 pesticides and factors of use and exposure, including frequency and duration, duration of residence on a farm, specific farm tasks performed, household practices of pesticides, and end-stage renal disease in the wives of pesticide applicators (n = 31,142). End-stage renal disease was higher in women who were obese, who used nonsteroidal anti-inflammatory drugs, or who had diabetes and hypertension. There was a protective effect associated with the personal use of any pesticide (hazard ratio [HR] = 0.42, 95% CI 0.28–0.64), but among women who personally mixed or applied pesticides, positive associations were observed only for the highest category of lifetime exposure days (HR = 4.22, 95% CI 1.26–14.20), although the estimate was imprecise. Wives with end-stage renal disease who reported having direct exposure to phenoxy herbicides (n = 9; HR = 1.10, 95% CI 0.50–2.39) or 2,4-D (n = 9; HR = 1.11, 95% CI 0.51–2.41) were found not to be at an elevated risk of the disease. Among wives who reported no direct use of pesticides but whose husbands used phenoxy herbicides (n = 47; HR = 0.86, 95% CI 0.49–1.51), 2,4-D (n = 45; HR = 0.82, 95% CI 0.47–1.43), 2,4,5-T (n = 12; HR = 0.69, 95% CI 0.36–1.32), or 2,4,5-TP (n = 5; HR = 0.88, 95% CI 0.36–2.15), the risks of end-stage renal disease were all decreased.
A third occupational exposure study was identified. 't Mannetje et al. (2018) conducted a morbidity survey among a subset of workers who were employed at the New Plymouth, New Zealand, phenoxy herbicide production plant for at least 1 month between 1969 and 1984. The plant produced 2,4,5-T, and the workers were potentially exposed to 2,4,5-T, the intermediates of TCP and other chlorophenols, as well as to TCDD. Workers had previously been recruited and examined as part of the international cohort of producers of phenoxy herbicides led by IARC (Kogevinas et al., 1997); see Chapter 5 for more details on the IARC cohort and the New Zealand phenoxy producers. This study extends the follow-up period of these workers to approximately 30 years from their last 2,4,5-T production exposure. From the original cohort of 1,025 workers, 631
were living, had a current address in New Zealand, and were below 80 years of age on January 1, 2006. For the current follow-up, 430 of the 631 workers were randomly selected and invited to participate in the morbidity survey, of whom 245 (57%) participated. The survey was administered in 2007–2008 by face-to-face interview, and information was collected on demographic factors and health information, including doctor-diagnosed conditions and the year of diagnosis. A blood sample was also collected at that time and was analyzed for TCDD, lipids, thyroid hormones, and other substances. For 111 participants, a neurological examination was conducted. Associations between exposure and health outcomes were assessed using logistic regression models that controlled for age, gender, smoking, BMI, and ethnicity using two different methods of exposure: having worked in a TCDD-exposed job (based on occupational records) and having serum TCDD concentration ≥10 pg/g lipid (18%). Mean TCDD concentrations were 19 pg/g lipid in the 60 men directly involved in phenoxy/TCP production and 6 pg/g lipid in the 141 men and 43 women who worked in other parts of the plant. Compared with the people in the non-highly exposed jobs, the people who had ever worked in a highly exposed job at the plant were no more likely to have a doctor-diagnosed kidney function problem (n = 13; OR = 0.82, 95% CI 0.34–1.99). When compared by serum TCDD concentration, no difference in the risk of kidney function problems was found for workers in the high- versus low-exposure groups (n = 5; OR = 1.03, 95% CI 0.32–3.33).
Using data from the 1999–2004 cycles of NHANES, Everett and Thompson (2016) evaluated the association of blood levels of three chlorinated dibenzop-dioxins, one chlorinated dibenzofuran, and four dioxin-like PCBs with nephropathy (microalbuminuria or macroalbuminuria) among 1,505 adolescents and young adults (12–30 years of age) with normal glycohemoglobin (A1c < 5.7%). In logistic regression models 1,2,3,6,7,8-hexachlorodibenzo-p-dioxin (OR = 51.1, 95% CI 4.1–641.6), PCB 126 (OR = 8.9, 95% CI 2.0–39.7), PCB 169 (OR = 9.4, 95% CI 1.02–87.6), and PCB 156 (OR = 17.9, 95% CI 2.1–152.6) were associated with nephropathy (OR = 7.1, 95% CI 1.8–28.1) when one or more of these four dioxin-like chemicals was elevated; however, the effect estimates are quite imprecise. The effect was driven by females (OR = 17.4, 95% CI 3.4–88.6), as among males there were no cases of nephropathy when one or more of the four dioxin-like chemicals were elevated. These results were verified by TEQs; TEQ8 ≥ 50.12 fg/g was associated with nephropathy among females (OR = 11.9, 95% CI 1.6–87.2) but not males. Thus, in a cross-sectional study, dioxin-like chemicals were associated with nephropathy among young females, but not males, though reverse causality cannot be excluded, and the effect estimates were very imprecise.
Two analyses using data from the cross-sectional, community-based study in the Annan District of Tainan City, Taiwan, where a former PCP factory had operated, and had released PCDD/Fs into the surrounding area, were identified that examined outcomes related to kidney disease (J. W. Chang et al., 2013; C. Y. Huang et al., 2016). As described in Chapter 5, people who were 18 years of age and older and who were residents of the exposure area were asked to participate in the study. Health examinations were performed on each participating individual, and serum samples had been previously collected and measured for levels of dioxins by the Tainan City Bureau of Health. A self-administered questionnaire, which was administered at the same time as the examination, was used to collect demographic information and medical history. C. Y. Huang et al. (2016) examined chronic kidney disease, defined as having an estimated glomerular filtration rate ≤ 60 mL/min/1.73m2 or having been diagnosed by a physician. People diagnosed with congenital kidney disease, IgA nephropathy, post-infectious kidney disease, or medicine-induced kidney disease were excluded from the study. Of the 2,828 participating individuals, 1,427 had high dioxin levels (defined as > 20 pg WHO98-TEQDF/g lipid in the serum), and 156 had chronic kidney disease. High dioxin levels were associated with an increased prevalence of chronic kidney disease compared with low dioxin levels (10.9% versus 1.6%, respectively, p < 0.001). After adjustment for PCDD/Fs, gender, mercury, metabolic syndrome, age, fasting glucose, insulin, and uric acid, a high dioxin level was found to be significantly associated with chronic kidney disease (OR = 1.74, 95% CI 1.02–2.97). The strengths of this study include a large population, adjustments for age, fasting glucose, insulin, and uric acid, as well as serum measurements of exposure and a clear definition of chronic kidney disease. However, this study is limited by having had no follow-up of renal function measurements, the fact that the serum PCDD/Fs levels were collected over an extended period of time that ended about 3 years before the interview and health examination, unknown age at first exposure to PCDD/Fs, unknown duration of exposure, unknown cumulative exposure dose, the cross-sectional design, and a lack of additional data collection. Data on other potential confounders, such as waist circumference, dietary intake, and socioeconomic status, were not available.
Raines et al. (2014) examined agricultural behaviors and health outcomes via questionnaire in a Nicaraguan community. Of the 424 total participants, 151 reported an occupational history of agriculture. The pesticides that were reported by participants as commonly used included 2,4-D. Decreased glomerular filtration rate was found in 9.8% of the women and 41.9% of the men. Glomerular filtration rate was associated with cutting sugarcane during dry season (OR = 5.86, 95% CI 2.45–14.01) and sugarcane chewing (OR = 3.24, 95% CI 1.39–7.58). Glomerular filtration rate was also associated with non-deliberate pesticide inhalation
(OR = 3.31, 95% CI 1.32–8.31). This study is limited by its lack of exposure validation through serum or other measures.
Other Identified Studies
Two other studies of kidney and urinary disorders were identified, but both were limited by a lack of exposure specificity (Orantes et al., 2015; Ruder et al., 2014). In a study of women in agricultural communities in El Salvador (Orantes et al., 2015), the authors examined exposures to various agrochemicals, which may have included organophosphate insecticides as well as phenoxy herbicides (2,4,-D, hedonal), but the results were not stratified by chemical exposures, and thus the study was not considered further.
A third study was also identified, but instead of being limited by exposure specificity, it was limited by the fact that the outcomes examined were not diagnosed health outcomes but rather indicators of biologic effects. In a separate analysis of people residing near the former PCP factory in Taiwan, J. W. Chang et al. (2013) evaluated associations between PCDD/Fs and the risk of hyperuricemia (too much uric acid in the blood) in a subset of healthy subjects from the community health study (n = 1,531). Serum levels of 17 2,3,7,8-substituted PCDD/Fs were measured, and associations were tested between the serum TEQDF-2005 (total PCDD/Fs 2005 WHO TEQ) and various factors, including uric acid, glomerular filtration rates, and hyperuricemia risk. Hyperuricemia is a measure of disturbed metabolism, not a health outcome, and therefore this study was not considered relevant to the committee’s task.
Studies of mice, rats, goldfish, and zebrafish have documented kidney toxicity from TCDD exposure. Studies of 2,4-D in goldfish and TCDD in rats report oxidative stress in kidneys (Matviishyn et al., 2014; Palaniswamy et al., 2014). Q. Liu et al. (2014) investigated the effects of developmental exposure to TCDD in zebrafish and found kidney lesions and the dysregulation of the genes involved in renal necrosis and cell death as well as decreased hematopoietic cells in the kidney marrow. Aida-Yasuoka et al. (2014) reported a chance finding that C57BL/6J mouse pups are more susceptible to TCDD-induced hydronephrosis than BALB/cA mice; in pups exposed to TCDD on postnatal days 1–7, the prevalence of hydronephrosis was 64% in the C57BL/6J pups and 0% in BALB/cA pups. In both strains of mice, the Ahr receptors are highly responsive to TCDD; however, genetic differences were found in expression of renal m-Prostaglandin E synthase-1 and early growth response 1 (Egr-1). In a study of mice, Bu et al. (2017) found that Ahr activation up-regulates glucose transporter 9 (Glut9), which plays a role in maintaining uric acid homeostasis.
Since Update 2014, seven studies have been reviewed for kidney disease and urinary disorder outcomes related to exposure to the COIs. A hospitalization study of New Zealand Vietnam veterans found that chronic renal failure risk was statistically significantly increased among the veterans compared with the standardized general population of New Zealand, but there was no difference in the prevalence of other kidney or urinary outcomes (Cox et al., 2015); however, there was no exposure validation, some of the conditions (such as urinary tract infections) do not typically require hospitalization, and potentially important risk factors were not adjusted for in the analysis. In a 30-year follow-up study of New Zealand workers in a plant that produced 2,4,5-T, having a doctor-diagnosed kidney function problem was not different for workers in high- versus low-exposure groups, based on reported job and duties in the plant or on serum TCDD levels. Two analyses of end-stage renal disease in the AHS (Lebov et al., 2015, 2016) found no statistically significant associations with the phenoxy herbicides 2,4-D or 2,4,5-T among the male pesticide applicators or their wives. In an analysis of NHANES, Everett and Thompson (2016) evaluated the association of the blood levels of three chlorinated dibenzo-p-dioxins, one chlorinated dibenzofuran, and four dioxin-like PCBs with nephropathy among 1,505 adolescents and young adults and found that dioxin-like chemicals were associated with nephropathy among young females, but not males, although the effect estimates were very imprecise. An environmental exposure study of Taiwanese residents living in close proximity to a former PCP-producing factory found that those who had high serum dioxin levels had a statistically significantly elevated risk of chronic kidney disease (Huang et al., 2016). A second analysis of this population was identified (Chang et al., 2013), but because it examined hyperuricemia, which is a measure of disturbed metabolism and not a recognized health outcome, it was not considered as part of the evidence base of chronic kidney conditions related to exposure to the COIs. A cross-sectional study of agricultural behaviors, including the use of 2,4-D, and health outcomes in a Nicaraguan community (Raines et al., 2014) found a decreased glomerular filtration rate to be associated with cutting sugarcane, sugarcane chewing, and non-deliberate pesticide inhalation, but no serum or other objective measure of exposure were collected. Studies of mice, rats, goldfish, and zebrafish document kidney toxicity from TCDD exposure, but such outcomes do not present a consistent mechanism for the kidney dysfunction diseases in humans. Epidemiology studies concerning exposure to the COIs and kidney diseases were not reported prior to Update 2014. However, the new epidemiologic findings reviewed do not present a coherent pattern of an association between exposure to the COIs and kidney or urinary disorders.
After reviewing the new evidence of exposure to the COIs and non-malignant kidney and urinary outcomes, the committee concludes that there remains inadequate or insufficient evidence of an association between exposure to the COIs and non-malignant kidney or urinary disorders.
This section discusses a variety of conditions related to endocrine function, excluding diabetes and other pancreatic disorders, which were discussed in Chapter 10: Cardiovascular and Metabolic Outcomes. In particular, clinical disruptions of thyroid function are grouped as ICD-9 240–246 or as ICD-10 E00–E07, E20–21, while the remaining endocrine disorders are grouped as ICD-9 252–259 or as ICD-10 E22–E35. Thyroid homeostasis in humans was first addressed with respect to the COIs by the committee for Update 2002.
The thyroid secretes the hormones thyroxine (T4) and triiodothyronine (T3), which stimulate and help to regulate metabolism throughout the body. The thyroid also secretes calcitonin, a hormone that controls calcium concentration in the blood and the storage of calcium in bones. Secretion of T4 and T3 is under the control of thyroid-stimulating hormone (TSH), which is secreted by the anterior pituitary. Iodine operates in thyroid physiology both as a constituent of thyroid hormones and as a regulator of glandular function. Concentrations of those circulating hormones are regulated primarily by a negative-feedback pathway that involves three organs: the thyroid, the pituitary, and the hypothalamus. In the hypothalamus–pituitary–thyroid feedback scheme, the hypothalamus releases thyrotropin-releasing hormone (TRH), which stimulates the pituitary to produce TSH, which in turn triggers the thyroid to produce T4 and T3. Cells in the hypothalamus and pituitary respond to concentrations of circulating T4 and T3. When T4 and T3 are low, the pituitary is stimulated to deliver more TSH to the thyroid, which increases T4 and T3 output. When circulating T4 and T3 are high, it triggers a reduction in the output of TRH and TSH. The negative-feedback loop maintains hormone homeostasis.
A disruption of thyroid homeostasis can be stimulatory (hyperthyroidism) or suppressive (hypothyroidism). Both conditions are diagnosed on the basis of blood concentrations of thyroid hormones, TSH, and other proteins (antithyroid antibodies). The prevalence of thyroid dysfunction in adults in the general population ranges from 1% to 10%, depending on the group, the testing setting, sex, age, the method of assessment, and the presence of conditions that affect thyroid function. People who have subclinical (biochemical) conditions may or may not show other signs or symptoms of thyroid dysfunction.
It is important to distinguish between potential effects on adults and effects that may occur during development. In adults, the thyroid is able to compensate, within reasonable limits, for mild or moderate disruption (such as that caused by hyperplasia or goiter). In contrast, the fetus is highly sensitive to alterations in thyroid hormones, and alterations in thyroid homeostasis can hamper the development of many organ systems, including the nervous and reproductive systems; such findings are discussed in Chapter 8, which addresses the potential effects of Vietnam veterans’ exposure to herbicides on their offspring. Only observations on adults are considered here.
Conclusions from VAO and Previous Updates
Thyroid homeostasis in humans was first addressed with respect to the COIs by the committee for Update 2002. After consideration of several new studies and because of the consistent observations of exposures to the COIs being related to perturbations of thyroid function—and to clinical hypothyroidism in particular—the committee for Update 2014 considered the body of epidemiologic data, in combination with strong biologic plausibility, to represent limited or suggestive evidence of an association between exposure to the COIs and hypothyroidism. Additional endocrine effects have been observed in conjunction with exposure to the COIs in both humans and animals, but the evidence is inadequate or insufficient to establish an association with herbicide exposure for them.
An extensive assessment of endocrine function in clinical examinations, including a series of thyroid-function tests, failed to show systematic differences in thyroid function between Ranch Hands and comparison veterans (AFHS, 1991a). In analyzing individual TCDD readings obtained for subjects in the AFHS, however, Pavuk et al. (2003) found statistically significantly increased TSH measures from the 1985 and 1987 examinations in the high-exposure category and a significant increasing trend across the three TCDD categories during the 1982, 1985, 1987, and 1992 examinations. No increased risk of disorders of the thyroid was found among a study of Australian Vietnam veterans (O’Toole et al., 2009). Among Korean Vietnam veterans, two publications considered thyroid outcomes (Yi et al., 2014a,b). The first (Yi et al., 2014a) used claims data to report on the prevalence of disorders of the thyroid gland in 111,726 veterans and found an increased risk of thyroid conditions overall (ICD-10 E00–E07) with herbicide exposure, after adjustment for several factors. The pattern was very similar for both non-iodine-deficiency hypothyroidism (ICD-10 E03) and for other nontoxic goiter (ICD-10 E04). The risk of thyroiditis (ICD-10 E06) overall was not found to be significantly associated with herbicide exposure, but the strongest endocrine-related results were for autoimmune thyroiditis (ICD-10 E06.3), a subcategory of thyroiditis. No difference in the risk of hyperthyroidism (ICD-10 E05) was found between high- and low-exposure groups. Comparing the high-exposure group with the low-exposure group based on EOI scores, an elevated
risk was observed for pituitary hypofunction (p = 0.011), while the risk of pituitary hyperfunction was not different between high- and low-exposure groups. The risk of hyperaldosteronism was also not elevated. A mortality analysis of the Korean Vietnam veteran cohort of 180,639 male veterans did not find any association between herbicide exposure and deaths from endocrine diseases when the cohort was analyzed as a group (Yi et al., 2014b).
Few studies of thyroid function have been conducted among occupational cohorts exposed to the COIs. Calvert et al. (1999) provided evidence of higher adjusted mean free-T4 concentrations in TCDD-exposed workers in the NIOSH Cross-Sectional Medical Study, but there was no dose–response relationship with serum TCDD. Bloom et al. (2006) found indications of an inverse relationship between the sum of dioxin-like chemicals and the concentration of free T4 in anglers in New York State, but no association between the sum of dioxin-like chemicals and TSH or T3. An analysis of the AHS that was restricted to male private pesticide applicators examined self-reported physician-diagnosed hyperthyroidism, hypothyroidism, and other thyroid conditions and the use of 50 specific agents, including 2,4-D, 2,4,5-T, and 2-(2,4,5-triphenoxy) proprionic acid (2,4,5-TP) (Goldner et al., 2013). After adjusting for age, education, and BMI, the odds of hypothyroidism for ever-use versus never-use were significantly elevated for 2,4-D, for 2,4,5-T, and for 2,4,5-TP. In comparison with those who had never used 2,4-D, an increased risk of hypothyroidism was seen in both those who had used 2,4-D for more than the median number of days and those whose days of 2,4-D use were fewer than the median (p-trend = 0.025). The use of 2,4,5-TP was found to be associated with a decreased risk of hyperthyroidism. None of the phenoxy herbicides were found to be related to having histories of other thyroid diseases. In an Italian study that compared urban and rural workers, Ciarrocca et al. (2012) found that the workers’ urinary arsenic levels differed by a factor of between 2 and 4, and urinary arsenic was positively correlated with serum TSH and thyroglobulin and negatively with free T3 and T4.
The results from several environmental studies have also been reviewed by update committees, and no evidence was found of effects on thyroid function or disease in women exposed to pesticides (Chevrier et al., 2008) or, among women in the AHS, exposed to phenoxy herbicides (Goldner et al., 2010). Using data from NHANES III (1988–1994), Schreinemachers (2010) did not find associations between recent exposure to 2,4-D and T4 or TSH concentrations. An analysis of 1999–2002 NHANES data (Turyk et al., 2007) found total T4 to have a weak inverse relationship with serum TEQs; the effect was somewhat stronger in people over 60 years old and in women.
Among women exposed in the Seveso incident, a significant inverse association was found between serum concentrations of TCDD over a 20-year period (1976–1996) with serum total T4, but not with TSH or free T3, which were measured in 1996. This association was stronger for women who were exposed before menarche than for women exposed after menarche. When thyroid
hormones were measured again in 2008 and compared with TCDD levels in 1976 and 1996, the association was no longer present (Chevrier et al., 2014). In a cross-sectional study of 2,264 Japanese men and women who had not been occupationally exposed to dioxins, self-reported thyroid disease (not otherwise specified) was not associated with serum levels of dioxin-like PCDD/Fs or dioxin-like PCBs or with total TEQs after adjusting for possible confounders (Nakamoto et al., 2013). Abdelouahab et al. (2008) described thyroid function in adult freshwater-fish consumers in Canada; dioxin-like congeners were associated with an increase in TSH and a decrease in T4 but below the threshold at which clinical symptoms would be present. Clear effects of dioxin-like chemicals on thyroid function were not apparent in Inuit adults (Dallaire et al., 2009) or in a cross-sectional study of a Chinese community exposed to an electronic-waste recycling plant (J. Zhang et al., 2010). Manh et al. (2013) and Kido et al. (2014) studied steroid hormone levels in the serum and saliva and dioxin concentrations in the breast milk of lactating Vietnamese women living in an “Agent Orange hot spot” (n = 51) or in an area with no suspected exposure (n = 58). Levels of cortisol and corticosterone in serum and saliva were higher in those women living in the hot spot area and were positively correlated with breast-milk dioxin concentrations. Trnovec et al. (2013) measured thyroid gland volume and free T4 in 320 adults from an organochlorine-contaminated area in Slovakia. Blood samples from these subjects produced readings above the limits of detection for all 7 dioxins, 8 of 10 furans, and all but 1 of the 12 PCBs on the 2005 WHO list of dioxin-like chemicals.
Update of the Epidemiologic Literature
One new epidemiological study of occupational exposures to the COIs and thyroid or endocrine effects has been identified for this update. Table 45, which can be found at www.nap.edu/catalog/25137, summarizes the results of studies related to thyroid homeostasis that have been reviewed in the VAO series.
't Mannetje et al. (2018) conducted a morbidity survey among a subset of workers who were employed at the New Plymouth, New Zealand, phenoxy herbicide production plant for at least 1 month between 1969 and 1984. The plant produced 2,4,5-T, and workers were potentially exposed to 2,4,5-T, the intermediates of TCP and other chlorophenols, and TCDD. Workers had previously been recruited and examined as part of the international cohort of producers of phenoxy herbicides led by IARC (Kogevinas et al., 1997); see Chapter 5 for more details on the IARC cohort and the New Zealand phenoxy producers. This study extends the follow-up period of these workers to approximately 30 years from their last 2,4,5-T production exposure. From the original cohort of 1,025 workers,
631 were living, had a current address in New Zealand, and were below 80 years of age on January 1, 2006. For the current follow-up, 430 of the 631 workers were randomly selected and invited to participate in the morbidity survey, of which 245 (57%) participated. The survey was administered in 2007–2008 by face-to-face interview and information was collected on demographic factors and health information, including doctor-diagnosed conditions and the year of diagnosis. A blood sample was also collected at that time and analyzed for TCDD, lipids, thyroid hormones, and other substances. Associations between exposure and health outcomes were assessed using logistic regression models that controlled for age, gender, smoking, BMI, and ethnicity using two different methods of exposure: having worked in a TCDD-exposed job (based on occupational records) and having serum TCDD concentration ≥10 pg/g lipid (18%). Mean TCDD concentrations were 19 pg/g lipid in the 60 men directly involved in phenoxy/TCP production and 6 pg/g lipid in the 141 men and 43 women who worked in other parts of the plant. Compared with the people in the non-highly exposed jobs, the people who had ever worked in a highly exposed job at the plant were no more likely to have a doctor-diagnosed thyroid disorder (n = 6; OR = 0.95, 95% CI 0.19–4.67). When compared by serum TCDD concentration, no difference in the risk of thyroid disorder was found for workers in the high- versus low-exposure groups (n = 5; OR = 4.00, 95% CI 0.76–21.0), but the estimate is very imprecise owing to the small number of cases.
Other Identified Studies
Three additional epidemiologic studies were identified that presented outcomes on endocrine and metabolic effects. A cross-sectional study of endocrine effects from the use of pesticides was conducted using a random sample of agricultural workers aged 18–69 years old in Brazil (Piccoli et al., 2016). Although questionnaires that collected detailed information on exposures were administered by trained interviewers to the participants and blood samples were collected from all participants to test cholinesterase activity, serum residues of organochlorine pesticides, and levels of free T4, total T3, and TSH, the researchers did not offer results for specific herbicides and organochlorine pesticides, thereby limiting the usefulness of this study to provide evidentiary weight regarding a potential association between exposure to the COIs and thyroid functioning. The second study was a cross-sectional study that examined whether the presence of metabolic syndrome altered serum concentrations of dioxin-like and non-dioxin-like PCBs under conditions of weight loss (Dirinck et al., 2016). Given the cross-sectional nature of the work, it is of limited usefulness in assessing the association of metabolic syndrome with dioxin-like chemicals.
A study by X. Sun et al. (2014) collected blood samples to determine the dioxin concentration and levels of nine steroid hormones from 48 men in an herbicide-sprayed hot spot and 36 men from a non-sprayed area. Participants
also completed a questionnaire about their health status. The levels of the steroid hormones, including testosterone, dehydroepiandrosterone, and estradiol, were measured and compared by exposure group. However, the differences in hormone levels are not surrogate measures of a health outcome, and, therefore, this study was not considered relevant to the committee’s charge.
The influence of TCDD on thyroid-hormone homeostasis has been measured in numerous animal studies, and TCDD exposure has been associated with changes in serum concentrations of T4, T3, and TSH. In most studies, TCDD exposure is associated with a hypothyroid state, including reduced circulating T3 and T4 and increased TSH, especially after chronic exposure. The reduction in circulating T4 concentrations is robust and has recently been proposed as a biomarker of the effect of dioxin-like chemicals (J. M. Yang et al., 2010). Female rats exposed chronically to TCDD showed follicular-cell hyperplasia and hypertrophy of thyroid follicles, which were consistent with an overstimulation of the thyroid by TSH (TSH increases as a homeostatic response to low T4 levels) (Yoshizawa et al., 2010). TCDD enhances the metabolism of thyroid hormones primarily through an AHR-dependent induction of glucuronyl transferase activity (Gessner et al., 2012; Y. Kato et al., 2010; Martin et al., 2012; Nishimura et al., 2005). An enhanced accumulation of T4 in hepatic tissue of TCDD-treated mice may also contribute to the reduction in circulating T4 (Y. Kato et al., 2010).
The possibility that arsenic could act as an endocrine disruptor on thyroid hormone–mediated processes has been proposed on the basis of cell-culture studies and experiments with the ex vivo amphibian tail metamorphosis assay (Davey et al., 2008). In guinea pigs that were fed diets containing 50 ppm arsenic as sodium arsenite or arsenic trioxide for 11 weeks, serum (total) T3 and T4 were reduced compared to controls by about 20–25% and 33%, respectively (Mohanta et al., 2014). Recent data in zebrafish show that arsenite exposure increases thyroxine levels and alters the expression of genes in the hypothalamic–pituitary–thyroid axis, including thyroid receptors α and β, TSH, and corticotropin-releasing hormone (H. J. Sun et al., 2015). These data raise the possibility that cacodylic acid may also disrupt thyroid homeostasis, but there are no published epidemiologic studies that have addressed this.
Numerous animal experiments and several epidemiologic studies have shown that TCDD and dioxin-like chemicals exert some influence on thyroid homeostasis, with the findings being most consistently indicative of hypothyroidism (Boas et al., 2006, 2012; Chevrier et al., 2014). The underlying molecular mechanisms resulting in these effects on thyroid hormone and TSH concentrations in humans,
however, are not yet fully characterized. In addition, there are some data to suggest the possibility that arsenic-based herbicides may also affect thyroid function.
Several previous studies of populations environmentally exposed to PCBs found some combination of elevated TSH concentrations and depressed T4 and T3 levels (Bloom et al., 2006; Hagmar et al., 2001a; Persky et al., 2001; Schell et al., 2004), although some (Hagmar et al., 2001b; Sala et al., 2001) found no significant effect. For example, Pavuk et al. (2003) reported a trend of higher TCDD serum concentrations being associated with increasing concentrations of TSH that was not accompanied by changes in circulating T4 or T3 (which would be interpreted as subclinical hypothyroidism) in participants in the AFHS. This finding in U.S. Vietnam veterans is complemented by the results from the Korean Veterans Health Study (Yi et al., 2014a), which found evidence of an increased occurrence of clinical hypothyroid disease, possibly associated with autoimmune hypothyroidism, in association with higher estimated potential herbicide exposure. In addition, the report from the AHS of increased physician-diagnosed hypothyroidism in herbicide applicators with phenoxy herbicide exposure (Goldner et al., 2013) supports the notion that this association is real. Results from the Korean Veterans Health Study suggest that adrenal and possibly pituitary function may also be affected by exposure to dioxin-like chemicals.
For the current update, in vitro and animal data continue to support previous findings that the COIs can alter thyroid homeostasis. There is additional new data (Sun et al., 2015) that raise the possibility that cacodylic acid might similarly disrupt the normal action of the thyroid. There is little additional epidemiologic data available regarding the association of the COIs with thyroid disease or other endocrine function. A follow-up of phenoxy herbicide producers in New Zealand did not find any difference in thyroid disorders between high- and low-exposure groups ('t Mannetje et al., 2018).
On the basis of the new evidence and that reviewed in previous VAO reports, the committee concludes that there is limited or suggestive evidence of an association between exposure to at least one of the COIs and hypothyroidism. There is inadequate or insufficient evidence for disruption of thyroid homeostasis or other endocrine disorders.
In previous VAO reports, skin disorders such as chloracne were mentioned, but they were not consistently included as independent outcomes. However, due to the nature of new published epidemiologic literature, the committee has reviewed several studies of exposure to the COIs that have included skin conditions, and thus the committee has decided to include them as a distinct outcome in this chapter.
Chloracne is a skin disease that is characteristic of high levels of exposure to TCDD and other diaromatic organochlorine chemicals. It is one of the few outcomes in humans that are consistently associated with such exposure, and it is a well-validated indicator of high-dose exposure to TCDD and related chemicals (Sweeney et al., 1997/1998). Chloracne shares some pathologic processes (such as the occlusion of the orifice of the sebaceous follicle) with more common forms of acne (such as acne vulgaris), but it can be differentiated by the presence of epidermoid inclusion cysts, which are caused by the proliferation and hyperkeratinization (horn-like cornification) of the epidermis and sebaceous gland epithelium. Although chloracne is typically distributed over the eyes, ears, and neck, it can also occur on the trunk, genitalia, and buttocks of chemical-industry workers exposed to TCDD (Neuberger et al., 1998).
If chloracne occurs, it appears within a few months after the chemical exposure, not after a long latent period; therefore, new cases of chloracne among Vietnam veterans would not be the result of exposure during the Vietnam War. Although it is resistant to acne treatments, it usually regresses.
The chronic skin conditions considered include skin infections, nuclear buds, karyolysis or karyorrhexis, comedones, scar formation, and skin pigmentation.
Conclusions from VAO and Previous Updates
The committee responsible for VAO (IOM, 1994) determined that there was sufficient evidence of an association between exposure to at least one COI (TCDD) and chloracne. Additional information available to the committees responsible for Update 1996 (IOM, 1996), Update 1998 (IOM, 1999), Update 2000 (IOM, 2001), Update 2002 (IOM, 2003c), Update 2004 (IOM, 2005), and Update 2006 (IOM, 2007) maintained that conclusion. Even in the absence of a full understanding of the cellular and molecular mechanisms that lead to the disease, several notable reviews (Panteleyev and Bickers, 2006; Sweeney and Mocarelli, 2000) have deemed the clinical and epidemiologic evidence of dioxin-induced chloracne to be strong. The occupational epidemiologic literature has many examples of chloracne in workers after reported industrial exposures (Beck et al., 1989; Bond et al., 1987, 1989a,b; Cook et al., 1980; Goldman, 1972; May, 1973, 1982; Oliver, 1975; Pazderova-Vejlupkova et al., 1981; Poland et al., 1971; Suskind and Hertzberg, 1984; Suskind et al., 1953; Zober et al., 1990). With relative risk estimates as high as 5.5 in exposed workers compared with referent non-exposed workers, Bond et al. (1989a) identified a dose–response relationship between probable exposure to TCDD and chloracne. Not everyone who is exposed to relatively high doses develops chloracne, and some with lower exposure may demonstrate the condition (Beck et al., 1989).
Almost 200 cases of chloracne were recorded among those residing in the vicinity of the accidental industrial release of dioxin in Seveso, Italy; most cases were in children and in those who lived in the highest-exposure zone, and most
Exposures of Vietnam veterans were substantially lower than those observed in occupational studies and in environmental disasters, such as in Seveso. The long period since the putative exposure has imposed methodologic limitations on the studies of Vietnam cohorts for chloracne. Nonetheless, the Vietnam Experience Study (CDC, 1988a) found that chloracne-like lesions were self-reported more often by Vietnam veterans than by Vietnam-era veterans (OR = 1.4; 95% CI 0.7–2.9). No excess risk was found in Vietnam versus era veterans among subjects who had dermatologic examinations (OR = 0.9, 95% CI 0.7–1.2 for comedones and acne from lesions; OR = 1.2, 95% CI 0.9–1.7 for hyperpigmentation). Compared with matched, non-exposed controls, Ranch Hand personnel reported a significant excess of acne (OR = 1.6, 95% CI 1.1–2.1) (Wolfe et al., 1990), but no cases of chloracne or post-inflammatory scars were found on physical examination 20 years after the potential herbicide exposure (AFHS, 1991b).
Update of Epidemiologic Literature
Three new studies of exposure to the COIs and chronic skin conditions have been identified since Update 2014. However, each study examined different outcomes, making comparisons among the studies difficult.
Cox et al. (2015) used hospital discharge records from 1988 to 2009 to identify health conditions that affected 2,783 male New Zealand veterans who had served in Vietnam during 1964–1972. Age-specific hospitalization rates were calculated using the total number of annual hospitalizations published by the Ministry of Health and the average annual resident population. Standardized hospitalization rates and 99% CIs were calculated for the veteran cohort and for the general population for several noncancerous conditions, and an SHR was calculated using the two rates. This analysis found that hospitalization for skin infections (n = 140; SHR = 1.22, 99% CI 1.15–1.21) was greater in Vietnam veterans than what would be expected based on same-age New Zealand males. However, there was no difference in hospitalizations for other skin diseases (n = 60; SHR = 1.17, 99% CI 0.87–1.46). This analysis was restricted to the first hospitalization for each cause in order to account for chronic disease. This analysis did not include information on or control for lifestyle factors or ethnicity. Exposure was not validated through serum measurements and was assumed based on deployment to Vietnam.
't Mannetje et al. (2018) conducted a morbidity survey among a subset of workers who were employed at the New Plymouth, New Zealand, phenoxy herbicide production plant for at least 1 month between 1969 and 1984. The plant produced 2,4,5-T, and workers were potentially exposed to 2,4,5-T, the intermediates of TCP and other chlorophenols, and TCDD. Workers had previously been recruited and examined as part of the international cohort of producers of phenoxy herbicides led by IARC (Kogevinas et al., 1997); see Chapter 5 for more details on the IARC cohort and the New Zealand phenoxy producers. This study extends the follow-up period of these workers to approximately 30 years from their last 2,4,5-T production exposure. From the original cohort of 1,025 workers, 631 were living, had a current address in New Zealand, and were below 80 years of age on January 1, 2006. For the current follow-up, 430 of the 631 workers were randomly selected and invited to participate in the morbidity survey, of which 245 (57%) participated. The survey was administered in 2007–2008 by face-to-face interview, and information was collected on demographic factors and health information, including doctor-diagnosed conditions and the year of diagnosis. A blood sample was also collected at that time and analyzed for TCDD, lipids, thyroid hormones, and other substances. For 111 participants, a neurological examination was conducted. Associations between exposure and health outcomes were assessed using logistic regression models that controlled for age, gender, smoking, BMI, and ethnicity using two different methods of exposure: having worked in a TCDD-exposed job (based on occupational records) and having serum TCDD concentration ≥10 pg/g lipid (18%). The mean TCDD concentrations were 19 pg/g lipid in the 60 men directly involved in phenoxy/TCP production and 6 pg/g lipid in the 141 men and 43 women who worked in other parts of the plant. Compared with the 124 people in the non-highly exposed jobs, the 121 people who had ever worked in a highly exposed job were no more likely to have doctor-diagnosed eczema (n = 17; OR = 0.57, 95% CI 0.27–1.21) or acne (n = 2; OR = 0.73, 95% CI 0.10–5.47). When compared by serum TCDD concentration ≥10 pg/g lipid, no difference in eczema (n = 8; OR = 1.31, 95% CI 0.47–3.61) or acne (based on 1 case) was found.
In 1968, bran rice oil was accidentally contaminated with high concentrations of PCBs, dioxins, and dioxin-like chemicals in Western Japan (the Yusho accident). People who were poisoned were recruited into a cohort (i.e., the Yusho Group) and received regular medical follow-up. Mitoma et al. (2015) examined skin symptoms in the Yusho Group (n = 352) who underwent dermatological check-ups in 2012. PCB and dioxin serum concentrations were still elevated in the Yusho Group compared to the general Japanese population, and skin diseases
were still very common. Furthermore, black comedones and scar formation were significantly correlated with blood 2,3,4,7,8-PeCDF level, and black comedones (n = 115; 32.7% of all patients), scar formation (n = 111; 31.5% of all patients), and skin pigmentation (n = 57; 16.2% of all patients) were significantly positively associated with the total serum PCB concentration. These results show that the prevalence of skin conditions such as black comedones, scar formation, and skin pigmentation are correlated with exposure to PCBs, dioxins, and dioxin-like chemicals, and that chloracne was persistent in this population 44 years after the acute ingestion of dioxins and dioxin-like chemicals.
Other Identified Studies
Four additional studies that reported skin conditions were identified, but each lacked the necessary exposure specificity to be considered further. A second study of long-term health effects from the Yusho accident (Akahane et al., 2017) examined the prevalence of self-reported conditions including diseases of the skin compared with an age-, sex-, and residential area–matched group. Because no TEQs or other quantification of relevant exposures was presented, the study was not considered further. Butinof et al. (2015) conducted a cross-sectional study of male pesticide applicators (n = 880) who were issued applicator licenses by the Agriculture, Livestock, and Food Ministry of Argentina between 2007 and 2010. All participants completed a self-administered questionnaire that was adapted from the U.S. AHS. The average worker was exposed to 11 different chemicals, and no pesticide-specific exposure assessment was conducted. Ruder et al. (2014) examined skin conditions such as chloracne and hyperpigmentation in a U.S. cohort of capacitor manufacturers exposed to dioxin-like and non-dioxin-like PCBs, but it was limited by a lack of exposure specificity. C.Y. Yang et al. (2015) evaluated 240 Yucheng, Taiwan, subjects aged 9–65 years who were exposed to PCB-contaminated rice oil in 1979 and who completed a health-related quality-of-life questionnaire about 30 years following the exposure. Exposure was not verified by serum measurements or other means.
A fifth study (Carbajal-Lopez et al., 2016) involved a cross-sectional study that examined biomarkers of DNA damage in buccal cells in male agricultural workers in Mexico who were exposed to pesticides that likely included 2,4-D and dicamba (n = 111) and compared them with males who had no occupational history of pesticide exposure (n = 60). Although significantly (p < 0.0001) more evidence of DNA damage, as measured by comet tail length, micronuclei, binucleated cells, pycnosis, and condensed chromatin, was found among the agricultural workers who were exposed to pesticides than among the non-exposed workers, these differences are not tied to an observable health outcome, and the study was not considered further.
Previous updates have reported that chloracne-like skin lesions have been observed in several animal species in response to exposure to TCDD but not to purified phenoxy herbicides. Data accruing over the past several decades demonstrated that TCDD alters the differentiation of human keratinocytes, and more recent studies have illuminated how. TCDD accelerates the events associated with early differentiation and also obstructs the completion of differentiation (Bock, 2017a,b; Fontao et al., 2017; Geusau et al., 2005; Mandavia, 2015). Panteleyev and Bickers (2006) proposed that the major mechanism of TCDD induction of chloracne is activation of the stem cells in the basal layer of the skin to differentiate and then an inhibition of their ability to commit fully to a differentiated status. De Abrew et al. (2014) elucidated the role of matrix metalloproteinase-10 in the histopathological changes that typify chloracne. Work by Tauchi et al. (2005) and K. J. Smith et al. (2017) implicated additional inflammation-related mechanisms by which TCDD exposure may lead to chloracne. The data provide biologically plausible mechanisms for the induction of chloracne by TCDD.
No epidemiologic data in the past decade have refuted the conclusion of prior VAO committees that the evidence of an association between exposure to dioxin and chloracne is sufficient. Because TCDD-associated chloracne becomes evident within a few months after exposure, there is no risk of new cases long after service in Vietnam. Given the established relationship of an association between TCDD and chloracne and the long period that has elapsed since service in Vietnam, the present committee concludes that the emergence of persistent additional biologic or epidemiologic evidence that would merit review and deliberation by later VAO committees is unlikely. The formation of chloracne lesions and other skin disorders after the administration of TCDD has been observed in some species of laboratory animals. The newly identified epidemiologic studies (Cox et al., 2015; Mitoma et al., 2015; 't Mannetje et al., 2018) examined different outcomes and had varying exposure specificity, thereby limiting the comparisons that could be made. Although some differences between exposed and unexposed populations were noted, each study had several limitations that did not change the evidence of association with exposure to the COIs and chronic skin conditions.
On the basis of numerous epidemiologic studies of occupationally and environmentally exposed populations and supportive toxicologic information, previous VAO committees have consistently concluded that there is sufficient evidence
of an association between exposure to at least one COI and chloracne. After review of new literature, the committee concludes that there is inadequate or insufficient evidence of an association between exposure to the COIs and other chronic skin conditions.
This section discusses eye problems (ICD-9 360–379; ICD-10 H00–H59). The loss of vision is increasingly common with advanced age, and about one-sixth of people over 70 years old have substantial impairment, with men and women being similarly affected (NCHS, 2010). The most prevalent ocular problems in the current age range of Vietnam veterans are age-related macular degeneration, cataracts, glaucoma, and diabetic retinopathy. Ocular problems involving chemical agents most often arise from acute, direct contact with caustic or corrosive substances that may have permanent consequences. Ocular impairment arising from systemic exposure to toxic agents may be mediated by nerve damage. Cataracts can be induced by a chronic internal exposure of the lens to such chemicals as 2,4-dinitrophenol, corticosteroids, and thallium; glaucoma may be secondary to a toxic inflammation or may result from topical or systemic treatment with anti-inflammatory corticosteroids (Casarett and Doull, 1995).
Conclusions from VAO and Previous Updates
Update 2010 considered one study of Australian Vietnam veterans that found that the veterans had a higher prevalence of all the eye conditions assessed—cataracts, presbyopia, color blindness, and other diseases of the eye—than the Australian population (O’Toole et al., 2009). However, the committee noted a lack of information on exposure to the COIs and a lack of clinical confirmation of the eye conditions, and it had serious concerns about the possibility that recall bias had played a role in the findings. On the basis of that one study, the committee for Update 2010 concluded that there was inadequate or insufficient evidence to determine whether there is an association between exposure to the COIs and eye conditions. No new epidemiologic studies of exposure to the COIs and eye problems had been published for review in Update 2012 and Update 2014.
Update of Epidemiologic Literature
Only one new study of eye conditions was identified. Cox et al. (2015) used hospital discharge records from 1988 to 2009 to identify prevalent health conditions in 2,783 male New Zealand veterans who had served in Vietnam during 1964–1972. Age-specific hospitalization rates were calculated using the total number of annual hospitalizations published by the Ministry of Health and the
average annual resident population. Standardized hospitalization rates and 99% CIs were calculated for the veteran cohort and the general population, and an SHR was calculated using the two rates for two conditions related to the eyes. Neither the risk of hospitalization for cataracts (n = 99; SHR = 1.11, 99% CI 0.82–1.40) nor retinal disease (n = 31; SHR = 1.35, 99% CI 0.73–1.98) was different between the veterans and general New Zealand population. Cataract and retinal disease are not generally conditions that require hospitalization, and therefore, the estimated prevalence may be higher. Exposure was not validated through serum measurements and was assumed based on deployment to Vietnam.
There have been several recent reports of ocular activity associated with Ahr activation or TCDD exposure of rats (Sugamo et al., 2009), mice (Takeuchi et al., 2009), and human non-pigmented ciliary epithelial cells (Volotinen et al., 2009). Hu et al. (2013) reported that mice harboring the null allele at the Ahr locus developed macular age-related degeneration-like pathology.
Woeller et al. (2016) investigated whether AHR activation had therapeutic potential for thyroid eye disease, which includes myofibroblast accumulation and orbital scarring. TGF-β induces myofibroblast formation, and AHR influences TGF-β signaling. Using orbital fibroblasts cultured from thyroid eye disease patients, Woeller et al. found that exposure to AHR ligands 6-formylindolo(3,2-b)carbazole2-(1′H-indole-3′-carbonyl)-thiazole-4-carboxylic acid methyl ester (FICZ and ITE) prevented TGF-β-induced myofibroblast formation.
Hospital discharge data from New Zealand Vietnam veterans (Cox et al., 2015) do not suggest any association between exposure to the COIs and retinal disease or cataracts. The data generated in vitro or in vivo in animal models reveal a quite modest possibility of ocular activity for the COIs; these are limited data that do not assist the committee in determining an association between exposure to the COIs and eye conditions.
Given the lack of additional evidence, the committee concludes that there is inadequate or insufficient evidence to determine whether there is an association between exposure to the COIs and eye conditions.
This section discusses osteoporosis, or decreased bone density (ICD-9 733.0–733.1; ICD-10 M80–M81), and other joint and connective tissue disease, excluding arthritis, which is discussed in Chapter 6: Immune System Disorders. Osteoporosis is a skeletal disorder characterized by a decrease in bone mineral density and a loss of the structural and biomechanical properties of the skeleton, which leads to an increased risk of fractures. Although there are no practical methods for assessing overall bone strength, bone mineral density correlates closely with skeletal load-bearing capacity and fracture risk (Lash et al., 2009). WHO has defined osteoporosis based on bone mineral density measurements. The diagnostic T-score derived by dual energy X-ray absorptiometry is the number of standard deviations from the mean bone mineral density for healthy women. In women, readings greater than −1.0 are normal, whereas osteopenia is defined by a T-score between −1.0 and −2.5, osteoporosis is defined by a T-score between −2.5 and −5.0, and severe osteoporosis corresponds to a T-score of −5.0 or lower. Diagnostic criteria have not been standardized for osteoporosis in men. Although men have much higher baseline bone mineral density than women, they seem to have a similar fracture risk for a given bone mineral density (Lash et al., 2009), so most authorities apply the same WHO ranges for T-scores calculated relative to normal young women.
Sex is an important risk factor for osteoporosis; about 56% of postmenopausal women have decreased bone mineral density, and 6% have osteoporosis (CDC, 2002). The effects of aging on bone loss in women are well known, but many health care providers and patients are less familiar with the prevalence and effects of bone changes in older men (Orwoll et al., 2010). Individual patients have genetic and acquired risks of osteoporosis, and the osteoporosis disease process can be without symptoms for decades. It is well known that hormones, vitamins, and pharmaceuticals can have adverse effects on bone and that drug-induced osteoporosis occurs primarily in postmenopausal women, but premenopausal women and men are also significantly affected. Glucocorticoids are the most common cause of drug-induced osteoporosis (Mazziotti et al., 2010). Other risk factors for the loss of bone mineral density include the use of long-acting benzodiazepine or anticonvulsant drugs, previous hyperthyroidism, excessive caffeine intake, and routinely standing for less than 4 hours per day (Lash et al., 2009).
Several studies have described a link between organochlorine exposure and effects on bone growth, most notably reports of infants exposed in utero to high concentration of PCBs and PCDFs who developed irregular calcifications of their skulls (Miller, 1985) and reports of accidental organochlorine poisoning that resulted in osteoporosis (Cripps et al., 1984; Gocmen et al., 1989). However, epidemiologic studies of the association between environmental exposures to organochlorine compounds and bone disorders have had inconsistent results.
Conclusions from VAO and Previous Updates
Update 2010 was the first report in the VAO series to review studies of the association between exposures to the COIs and a decrease in bone mineral density. Few studies have reported results in this category with enough exposure specificity to be fully informative. Hodgson et al. (2008) studied the relationship between environmental exposures and bone mineral density in a set of 325 members of the Osteoporosis Cadmium as a Risk Factor (OSCAR) cohort who were at least 60 years old. Forearm bone mineral density was measured, blood samples were analyzed for the five dioxin-like mono-ortho PCB congeners (PCB 105, 118, 156, 157, and 167), and TEQs were calculated. In men, PCB 118 was significantly negatively associated with bone mineral density, but there was no association between the TEQ for any of the five dioxin-like mono-ortho PCBs and bone mineral density. In women, PCB 118 and the TEQs for all five dioxin-like mono-ortho PCBs were positively associated with bone mineral density. When the risk of low bone mineral density was treated as a binary variable in an adjusted logistic model, there was a significant association with PCB 118 in men, but none of the measured compounds was predictive in women. A study of 350 women who were exposed to TCDD as a result of the chemical explosion in Seveso, Italy, examined the relationship of DEXA-assessed bone mineral density and TCDD serum levels (Eskenazi et al., 2014). The results suggested that TCDD levels were associated with some evidence of better bone structure in the 48 women for whom exposure occurred after peak bone mass, which is estimated to happen 2 years after menarche, but the findings did not support the hypothesis that postnatal TCDD exposure adversely affects adult bone health. Two cross-sectional studies that examined the association of exposure to dioxin-like chemicals and bone quality in residents of Canada’s northern regions, who are known to be exposed to these chemicals as a result of their diet of marine mammals and predatory fish (Paunescu et al., 2013a,b), found an increase in dioxin-like PCBs 105 and 108 to be negatively associated with a “stiffness index,” and neither total plasma dioxin-like chemicals levels nor any specific dioxin-like PCB level was associated with ultrasonography-assessed bone strength in Inuit women.
Update of the Epidemiologic Literature
Only one new study of bone conditions was identified. Cox et al. (2015) used hospital discharge records from 1988 to 2009 to identify health conditions that had affected 2,783 male New Zealand veterans who had served in Vietnam during 1964 to 1972. Age-specific hospitalization rates were calculated using the total number of annual hospitalizations published by the Ministry of Health and the average annual resident population. Standardized hospitalization rates and 99% CIs were calculated for the veteran cohort and the general population, and an SHR was calculated using the two rates for conditions related to other joint
diseases and other connective tissue diseases. An elevated risk of hospitalization for other joint diseases (n = 77; SHR = 1.34, 99% CI 1.04–1.64) and other connective tissue disease (n = 169; SHR = 1.25, 99% CI 1.06–1.44), neither of which is further defined, was found for veterans when compared with the standardized New Zealand population. Since these disease categories were not clearly defined, it is difficult to interpret the findings. The study is also limited because the exposures were not validated through serum measurements and were assumed based on deployment to Vietnam.
Animal studies suggest that TCDD may have some influence on bone formation and maintenance. Recent work from Herlin et al. (2013) showed that the exposure of adult mice to TCDD resulted in harder bone matrix, thinner cortical bone, mechanically weaker bones, and, most notably, increased trabecular bone volume fraction in Ahr knockout mice. It is known that TCDD can induce chondrocyte apoptosis in culture, which could be an initial event leading to the sort of cartilage degradation observed in arthritis (Yang and Lee, 2010). Lee and Yang (2012) recently demonstrated that this is mediated by reactive oxygen species. In addition, TCDD exposure via the dam’s milk impaired bone mineralization during postnatal development in mice because of a reduction in osteoblastic activity as a result of TCDD-induced up-regulation in the active form of vitamin D in serum (Nishimura et al., 2009). TCDD altered osteogenesis (bone formation) in an in vitro osteoblast model and led to alterations in the proteins associated with cytoskeleton organization and biogenesis, a decrease in the expression of calcium-binding proteins, and a decrease in osteoblast calcium deposition (Carpi et al., 2009). In adult rats, TCDD exposure reduced trabecular bone cross-sectional area, but it significantly increased total bone mineral density; it was further noted that TCDD decreased the expression of the bone-formation marker procollagen type I N-terminal propeptide and increased the expression of the bone-resorption marker carboxy-terminal collagen cross-link, suggesting a net loss of bone tissue (Lind et al., 2009). It is also known that exposure to polyaromatic hydrocarbons (such as those in tobacco smoke) can affect bone health, and some of these alterations have been shown to be mediated, at least in part, by the AHR. That implies that TCDD may alter or modify the effects of osteoblast formation and function as well as inhibit osteoclast formation and function (Kung et al., 2012; Yan et al., 2011). Iqbal et al. (2013) recently addressed this, studying genetically altered mice in order to understand the contributions of tobacco carcinogens and TCDD. In their work, mice in which the Ahr or CYP1A1, CYP1A2, and CYP1B1 genes were deleted displayed reduced resorption and high bone mass. In contrast, Ahr activation by administering benzo[a]pyrene to wild-type mice increased osteoclastogenesis and bone resorption.
The small amount of available epidemiologic information on the possible adverse effects of exposure to the COIs on bone structure has previously been based entirely on dioxin-like mono-ortho PCBs, which contribute a small percentage to total TEQs based on all dioxin-like PCBs. The recent data from New Zealand Vietnam veterans, while suggesting a slightly statistically elevated risk of “other” bone and joint conditions, has poorly described outcomes (with essentially no data on the exact bone conditions considered) and is further limited by the use of hospitalization rates for such outcomes, which are difficult to interpret. Biological data confirm that TCDD is active in bone metabolism, but the pattern of association of exposure to the COIs and subsequent disease is not consistent in the current literature.
There is inadequate or insufficient evidence of an association between exposure to the COIs and clinical or overt adverse effects of osteoporosis, loss of bone mineral density, or other bone conditions.