Chapter Three:
Possible Health Effects of Oil Fires

This chapter describes the possible health effects due to exposures to the pollutants of concern. The information and conclusions that follow are based on key primary sources, i.e., publications in peer-reviewed scientific journals. The information presented here has been restricted to the range of exposures of interest that are based on the maximum concentrations measured in the region as depicted in Table 2.11. The section describes what is known about the health effects of VOC, PAH, particulate matter, acidic aerosols, metals, criteria and photochemical pollutants, and long-term exposures to air pollution. The last section summarizes the results of several health studies related to Gulf War veterans conducted during and after the conflict. It also includes an overview of the symptoms reported by Gulf War veterans to complete the stand-alone report. Again, it in no way purports to be either a critical or comprehensive review.

Volatile Organic Compounds


Benzene is present, at low levels, in many plants and animals. Although natural sources such as volcanoes and forest fires emit small amounts of benzene, the major releases arise from the use of crude oil and gasoline, and from chemical-industry emissions. The most common sources of benzene exposure for humans are gasoline filling stations, tobacco smoke, and vehicle exhaust fumes. Because benzene is very volatile, the prevalent form of exposure is by inhalation, followed by ingestion of contaminated foods and water, and last by dermal contact, mainly with products containing benzene such as gasoline.

For the general population, benzene standards are set by the EPA for drinking water at 5 ppb (1 ppb = 3.20 µg/m3), with the ultimate goal of 0 ppb in drinking water and in lakes and rivers. NIOSH's recommended occupational TWA is 100 ppb, OSHA's TWA is 1000 ppb, and ACGIH's TLV is 500 ppb.

Benzene is ubiquitous in air, both in rural and urban areas as well as indoors. Table 3.1 lists benzene concentrations in several U.S. locations and at the Al Ahmadi hospital.

Table 3.1
Benzene Concentrations in Outdoor Air

Locations Mean Concentration Range Comments References
Denver, CO 24.5
Urban, summer 1986 EPA, 1987
73 km NW of Denver, CO 0.02-0.85 Rural, spring-fall 1982 Roberts, 1985
Manhattan, NY 10.5
Urban, summer 1986 EPA, 1987
Staten Island, NY 6.6
Urban, spring 1984 Singh, 1985
Elizabeth & Bayonne, NJ 3.0 max.
Urban, daytime, fall 1981 Wallace, 1985
Elizabeth & Bayonne, NJ 2.7 max.
Urban, at night, fall 1981 Wallace, 1985
Chicago, IL 20.7
Urban, summer 1986 EPA, 1987
St. Louis, MO 11.1
Urban, summer 1985 EPA, 1987
Stinson Beach, CA 0.38 � 0.39 Remote coastal, 1984 Wester, 1986
Al Ahmadi Hospital 4.2 Maximum USAEHA, 1994

Daily median benzene air concentrations for the period 1975-1985 are listed in Table 3.2. The outdoor air values include data from 300 cities in 42 states and the indoor air from 30 sites in 16 states (Shah, 1988).

Table 3.2
Daily Median Air Benzene Concentrations

Location Concentration
Remote 0.16
Rural 0.47
Suburban 1.8
Urban 1.8
Indoor air 1.8
Workplace aira 2.1

SOURCE: Shah, 1988.

aThese measurements were made prior to current restrictions on smoking in the workplace. Current levels are most likely lower.

Table 3.3

Benzene Concentrations in Air at Different Sites

Site Concentration Reference
Inside home w/ 0 smokers 2.2 Wallace, 1989a
Inside home w/ 1 or more smokers 3.3 Wallace, 1989a
Inside a smoke-filled bar 8.1-11.3 Brunnemann, 1989
Breath of smokers 4.7 Wallace, 1989b
Breath of nonsmokers 0.47- 0.63 Wallace, 1989b
Breath of smokers in urban area 6.8�3.0 Wester, 1986
Breath of nonsmokers in urban area 2.5 � 0.8 Wester, 1986
Breath of smokers in remote area 2.1 � 9.6 Wester, 1986
Breath of nonsmokers in remote area 1.8 � 0.2 Wester, 1986
In vehicle on NJ Turnpike 5.0 � 6.0 Lawryk, 1995
In vehicle in Lincoln Tunnel, NY-NJ 8.1 � 8.3 Lawryk, 1995
Municipal landfill 32 Wood, 1987
Kin-Buc Landfill, Edison, NJ (Superfund site) 59.5 Bennett, 1987
Love Canal basement, Niagara Falls, NY 162.8 Pellizzari, 1982

Personal exposures[1] to benzene tend to exceed the outdoor air concentrations. Data from the Total Exposure Assessment Methodology (TEAM) study (Wallace, 1989a) give a mean personal exposure of about 4.7 ppb, compared to a mean outdoor concentration of only 1.9 ppb. The same study also measured the median level of benzene inside homes without smokers and with one or more smokers. The results are included in Table 3.3. It is interesting to note that the personal level is higher than the level inside the home. A plausible explanation is that the personal value also includes exposures at other locations where daily activities of the study volunteer take him or her, i.e., in transit, inside a car, at work, etc.

Benzene intake during daily activities can be estimated. A smoker who consumes approximately two packs per day will have an additional daily intake of about 1200 µg of benzene (Fishbein, 1992). Assuming an urban concentration range of 2.8-20 ppb and an air intake of 20m3 per day, then the average air intake of benzene is 180-1300 µg per day. For a moving automobile with an average benzene concentration of 40 µg/m3 and an exposure duration of one hour per day, the benzene intake would be approximately 40 µg per day (Wallace, 1989a). Estimates of exposure from self-filling a car with gasoline and from evaporative emissions seeping into a home from automobiles in attached garages have been set at 150 µg per day (Wallace, 1989a).

Human studies show that inhalation exposure to benzene in the 1,000-ppb range (1 ppm = 1,000 ppb) from several months to several years reduces the number of all major blood cell types--erythrocytes (red blood cells), platelets, and leukocytes (white blood cells)--that are produced in the bone marrow. The next stage of severity is aplastic anemia, when the bone marrow ceases to function. Aplastic anemia can progress to acute myelogenous leukemia. Several occupational studies of workers exposed to low levels of benzene (approximately 25,000 ppb for nine years (Fishbeck, 1978) and 2,000-35,000 ppb (Townsend, 1978)) indicate a slight decrease in red blood count (RBC) at the end of the exposure period, but normal values years later. Significant decreases in white- and red-cell counts were recorded for workers exposed to 75,000 ppb for 10 years; for later years at lower exposures (15,000-20,000 ppb) their blood counts increased to normal values (Kipen, 1989). More severe effects, such as preleukemia or acute leukemia, were observed in workers exposed to 210,000-650,000 ppb for 1-15 years (Aksoy, 1978). Painters exposed to 3,000-7,000 ppb of benzene and other VOCs for 1-21 years showed increased serum immunoglobin (IgM) values and decreased values of IgG and IgA (Lange, 1973).

Multiple animal studies support the above observations (Rozen, 1984, 1985). Animal studies indicate a decrease in functional immune responses reflected in decreased resistance to infectious agents at benzene concentrations > 30,000 ppb for five days with recovery on the seventh day (Rosenthal, 1985).

The genotoxicity[2] of benzene has been studied extensively. Benzene and its metabolites seem to be genotoxic to humans, causing primarily chromosomal aberrations (Major, 1992; Yardley-Jones, 1990; Sasiadek, 1989).

Experimental data for animals and studies of humans indicate a link between a decrease in bone-marrow cellularity and the development of leukemia. Many cases of benzene-induced leukemia seem to have been preceded by aplastic anemia (Toft, 1982). Benzene is considered a human carcinogen by U.S. and international agencies; the EPA classifies it as a Class A variety. The EPA estimated a risk value of 2.7 x 10-2 for leukemia due to a total lifetime exposure of 1,000 ppb inhaled benzene for concentrations in air below 31 ppb (EPA, 1986).


Toluene, a very volatile liquid, occurs naturally in petroleum crude oil and in the tolu tree (hence its name). The most preponderant sources of toluene are evaporation from gasoline and release through car exhaust. Levels of toluene in the outdoor environment range from 0.26-7.8 ppb (1-30 µg/m3) in suburban and urban air (1 ppm = 3.75 µg/m3) (Chan, 1991). However, the maximum air concentrations of toluene are found indoors in the range of 9.9-18.5 ppb (38-71 µg/m3) (Wallace, 1991). This increased concentration arises from the use of household products such as paint, thinners, and glues; attached garages and cigarette smoking are also major household contributors.

The Clean Air Act Amendments of 1990 list toluene as a hazardous air pollutant. OSHA's TWA is 200 ppm (= 200,000 ppb = 750,000 µg/m3); NIOSH's TWA is 100 ppm (= 100,000 ppb = 375,000 µg/m3); and ACGIH's TLV is 50 ppm (50,000 ppb = 188,000 µg/m3).

Studies of humans and animals have demonstrated that toluene is readily absorbed via the lungs and the gastrointestinal tract; it accumulates in adipose tissues (EPA, 1984). Elimination of toluene is primarily (approximately 2/3) via urine as hippuric acid and is usually complete within 24 hours of exposure. Low and intermediate levels of exposure to toluene primarily affect the central nervous system as summarized in Table 3.4 (Benignus, 1981). Effects were reversible, even at high-exposure levels for long durations.

Table 3.4
Health Effects of Toluene Exposure in Humans

Concentration Duration Symptoms Reference
100 ppma 4 days @ 6 hours per day Headaches, dizziness, and eye irritation Andersen, 1983
600 ppm 8 hours All of the above plus euphoria, dilated pupils, convulsion and nausea Benignus, 1981
200-800 ppm chronic All of the above plus fatigue, muscular weakness, confusion, and accommodation disturbances Greenberg, 1997;
Boey, 1997
10,000-30,000 ppm Narcosis and death Echeverria, 1989

appm = 1,000 ppb.

Epidemiological studies revealed no significant increased risk for cancer among workers, but toluene may produce liver and kidney damage at high levels of exposure (Benignus, 1981).


Xylene occurs naturally in petroleum and coal tar, and is also formed during forest fires. It is a compound primarily used as a "safe" substitute for benzene, and is found in gasoline as part of the BTX component (benzene-toluene-xylene). Xylene has wide industrial use as a solvent and in the manufacture of synthetic agents.

Very little is known of its human health hazards compared to benzene and toluene, particularly chronic effects. Current animal data about whether xylene causes cancer are inconclusive. Epidemiological studies on cancer risks associated with toluene and xylene have to control for the known effects of benzene impurities (McMichael, 1988). Low levels (100-300 ppm) of inhaled xylene can cause eye, nose, and throat irritation, delayed response to visual stimuli, and reduced memory (Fishbein, 1985). NIOSH's and OSHA's TWA as well as ACGIH's TLV for m,o,p-xylene is 100 ppm. Exposure of workers in China to a mixture of toluene and xylene indicates that the effects of the combined toxicities are additive (Chen, 1994).

Polycyclic Aromatic Hydrocarbons

PAHs are produced by incomplete combustion of coal, oil, gas, forest vegetation or other organic substances. Although only a few PAHs have limited commercial use, they are found throughout the environment in air, water, and soil. Sources include vehicle exhaust, asphalt roads, coal tar, coal, and hazardous waste sites. Most of the health effects of individual PAHs have not been clearly identified. Although they are dissimilar, they tend to appear in groups rather than individually, depending on the process; thus, they are considered a group. The PAH group is also known as Coal Tar Pitch Volatiles or the fused polycyclic hydrocarbons that volatilize from the distillation residues of coal, petroleum, etc. It consists of the following compounds: The PAH's standards are set for coal-tar pitch volatiles determined as the cyclohexane-extractable fraction. NIOSH's TWA is 100 mg/m3; OSHA's TWA and ACGIH's TLV are 200 mg/m3.

Background levels of the PAH group in air in the United States are reported to be in the range of 20-1200 µg/m3 in rural areas and 150-19,300 µg/m3 in urban areas. Personal air concentrations of benzo(a)pyrene in Padua, Italy, were estimated in winter and summer; the means were 370 µg/m3 and 121 µg/m3 respectively (Minoia, 1997). Similar measurements were taken on the street and inside a city park in Copenhagen, Denmark; the results were 4400 µg/m3 and 1400 µg/m3 respectively (Nielsen, 1996). Background levels of PAHs in drinking water are in the range of 4-24 ng/L.

Humans are exposed to PAHs by choice of lifestyle and culture. For example, a study of commercial suntan oils based on mineral or vegetable oils that were analyzed for PAHs showed that all the samples contained benzo[a]pyrene together with other mutagenic, co-carcinogenic or noncarcinogenic PAHs, fluoranthene, benzo[k]fluoranthene, and anthracene. The total PAH content of the samples varied from 89-189 ng/g, while benzo[a]pyrene levels were in the 2-5 ng/g range. The results suggest that users of suntan oils may be exposed to low levels of potentially hazardous PAHs (Monarca, 1982).

In another study, the levels of 13 PAHs were determined in smoked fishery products from both modern smoking kilns with external smoke generation and from traditional smoking kilns. The average benzo(a)pyrene (BaP) concentration in all 35 samples from modern smoking kilns was 0.1 µg/kg (wet weight). The sum of other PAHs determined in the study (benz(a)anthracene, chrysene, benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(a)pyrene, dibenz(a,h)an-thracene and indeno(1,2,3-c,d)pyrene) was about 4.5 µg/kg (wet weight). The BaP levels of the 27 smoked fish samples from traditional kilns ranged from 0.2-4.1 µg/kg, with a mean value of 1.2 µg BaP/kg. The average concentration of the sum of the carcinogenic compounds was 9.0 µg/kg (Karl, 1996). Another dietary study from Italy calculated that the total dietary PAH intake was 3 µg per day per person, and the total intake of carcinogenic PAHs was 1.4 µg per day per person--high compared to the calculated inhalation of 0.370 µg per day in the most-polluted cities (Lodovici, 1995).

Scant information is available on the human health effects of specific PAH compounds. The limited health-effects data available arises from animal studies, where a wide range of effects have been found but only from exposure to extremely high doses of benzo(a)pyrene by ingestion, dermal contact, or prolonged inhalation (Sharma, 1997). The PAHs in these studies include anthra-cene, benzo(a)pyrene, benzo(b)fluoranthene, benzo(k)fluoranthene, chrysene, dibenz(a,h)anthracene, and indeno(1,2,3-cd)pyrene. There is no conclusive evidence that similar effects could occur in humans. However, the U.S. Department of Health and Human Services has determined that PAHs may reasonably be considered carcinogens.

Recent epidemiological studies report direct evidence of the carcinogenic effects of PAHs in occupationally exposed subjects. Risks of lung and bladder cancer were dose-dependent when PAHs were measured quantitatively against unexposed control groups. These findings suggest that the current threshold limit value of 200 µg/m3 of benzene-soluble matter, which indicates PAH exposure, may be too high; after 40 years of exposure, it gives a relative risk of 1.2-1.4 for lung cancer and 2.2 for bladder cancer (Mastrangelo, 1996). Studies indicate that when binary mixtures of some PAHs are administered, the yield of nuclear anomalies in the mouse gastrointestinal tract is less than expected by simple addition and closer to that expected by averaging the activities of the two PAHs comprising the mixture (Reddy, 1991).

Particulate Matter

Particulate matter (PM) is the generic term for a broad class of chemically and physically diverse substances that exist as discrete particles (droplets or solids) over a wide range of sizes. Particles originate from a variety of anthropogenic stationary and mobile sources as well as from natural sources. Particles may be emitted directly or formed in the atmosphere by transformations of gaseous emissions such as sulfur oxides (SOx), nitrogen oxides (NOx), and VOCs. The chemical and physical properties of PM vary greatly with time, region, meteorology, and source category, thus complicating the assessment of health effects.

Epidemiological studies on the health effects of particulate matter suggest significant short- and long-term toxicity at current ambient levels in the United States. So far, these effects seem to be less influenced by particle composition, inorganic versus organic, and nominal size than by gravimetric estimates of exposure. This appears to contradict the premises of conventional air-pollution toxicology, which is based on chemical-specific toxicity and the critical role of size in particle potency. The biological mechanism underlying the health effects found with PM10 in epidemiological studies is not well understood.

Time-series mortality studies suggest that an increase of 50 µg/m3 in the average 24-hour exposure to PM10 above the NAAQS is associated with an increased relative risk (RR)[3] that ranges between 1.025 and 1.05 in the general population, and an even higher RR in at-risk subpopulations, i.e., the elderly and those with pre-existing respiratory conditions (Pope, 1996; Schwartz, 1996a, 1996b). The range of PM10-mortality RR across studies may reflect the likely differences in PM10 composition as well as differences in PM10-averaging periods considered in the analyses. Recent analyses of the Harvard Six City Study that was conducted in six eastern U.S. cities found larger associations between excess mortality and fine particles (PM2.5), than with coarse particles (PM10-PM2.5) alone (Schwartz et al., 1996). Moreover, the correlation of excess mortality with coarse-mass particles becomes not significant, except in Steubenville, OH, where the coarse particles are probably predominantly from industrial combustion sources (RR = 1.053 per 25 µg/m3 in PM2.5).

Many studies have investigated the relationship between hospital admissions, outpatient visits, and emergency room visits for respiratory and heart diseases and PM10 in conjunction with other pollutants, e.g., O3, SO2, CO, NO2, H+. These studies have used data from many cities in the United States and Canada (Burnett, 1994; Schwartz, 1994a,b,c, 1995a, 1996a; Gordian, 1996). Chronic obstructive pulmonary disease (COPD), pneumonia, and nonspecific respiratory-disease hospitalizations show moderate but statistically significant RR in the range of 1.06-1.25 when there is an increase of 50 µg/m3 in PM10 or its equivalent. Although a substantial number of hospitalizations for respiratory-related illness occur in those older than 65, there are numerous hospitalizations for those under 65 as well.

Many of these studies also examine the effect of O3, and collectively they indicate that ambient O3 has a significant effect on hospital admission for respiratory causes, with RR ranging from 1.1 to 1.36 per 100 ppb. For a two-pollutant model (PM10 + O3), the RR range is 1.04 to 1.54 per 100 µg/m3, while individually the RR for O3 is slightly lower than for PM10 (Schwartz, 1995a). The PM10 and O3 effects appear to be independent of each other, with no reduction in the RR for one pollutant after control for the other. Also, there is a suggestion of an effect between PM10 and heart disease; there is none for O3.

Studies of acute respiratory illness include upper respiratory, lower respiratory, or cough in children and in adults. Studies of upper-respiratory illness do not show consistent results. Three studies show a RR of 1.2 (Pope, 1991, 1992; Hoek, 1993), and another study estimates it as 1.55 (Braun-Fahrländer, 1992). These inconsistencies could be attributed to the difference in populations.

Studies of lower-respiratory disease give RR ranging from 1.10 to 1.28 (Pope, 1991, 1992; Hoek, 1993; Schwartz, 1991a) and another study set it at 2.0 (Schwartz, 1994). Studies of cough were more consistent, with RR ranging from 0.98 to 1.51 (Pope, 1992; Hoek, 1993; Schwartz, 1994; Dusseldorp, 1994). These estimated RR have a larger scatter than the corresponding RR derived from the hospitalization data. This variability is probably due to the former studies' populations including several different subgroups, whereas the hospitalization studies tended to include more uniform populations.

Limited evidence suggests pulmonary function decrements are associated with chronic exposure to particulate matter indexed by various measures--Total Suspended Particulate (TSP), PM10, sulfates, etc. (Spektor, 1991; Ackerman-Liebrich, 1997; Raizenne, 1996). These cross-sectional studies require a very large sample size to detect differences because of the inherent person-to-person variability; therefore, lack of statistical significance cannot be construed as proof of null effect.

Ultrafine Particles

The ultrafine, or nucleation mode, particles have a median diameter of about 0.02 µm. They have an approximately six-order-of-magnitude higher number concentration and a hundred times higher surface-to-volume ratio than a 2.5 µm-diameter particle when inhaled at the same mass concentration. This means, for example, that for the same mass of particles the ultrafine particles will present a much larger surface where gases could adhere and be carried into the lungs via inhalation, potentially increasing toxicity because of the deeper penetration into the lung. Single ultrafine particles occur in the urban atmosphere in high number concentrations, 5x104-3x105 particles/cm3, but in low mass concentrations (Brand, 1991). Fortunately, single ultrafine particles are not very stable and eventually aggregate into larger particles, but new particles are always being produced from anthropogenic sources, e.g., gas-to-particle conversion, combustion processes, incinerator emissions, etc.

Although virtually no human studies exist on the health effects of ultrafine particles, some relevant issues have been examined. A recent study on the effect of ambient-particle size on lung function and symptoms in asthmatics suggests that the effects of the number of ultrafine particles (diameter < 0.1 µm) may be greater than those of the mass of fine particles (diameters of 0.1 to 2.5 µm) (Peters, 1997). An animal study that investigated the surface coating of ultrafine particles suggests that acute exposure to near-ambient concentrations of sulfuric acid under conditions that promote the formation of acid as a surface coating on respirable particles can induce an enhanced nonspecific airway hyperresponsiveness (Chen, 1992).

Acidic Aerosols

Several major pollution episodes have occurred in this century. Among these are the Meuse Valley, Belgium, in December 1930; in Donora, PA, in October 1948; and in London in December 1952 and December 1962. These specific historical episodes highlighted the major health effects of acidic aerosols on general populations. Unfortunately, no actual ambient air measurements during the first two episodes are available. In the Meuse Valley incident, although many pollutants existed in the atmosphere, the observed health effects were most strongly associated with sulfuric acid (Firket, 1936). More than 60 people died from this acidic fog, over ten times the normal rate, and hundreds suffered respiratory problems. Those especially affected were the elderly, asthmatics, and heart patients.

About 42 percent of the Donora population experienced deleterious effects from its three-day smog episode. Mild upper-respiratory tract symptoms were evenly distributed through all age groups and on average lasted for four days. More than half of those above 55 complained of dyspnea, the most common symptom. The observed health effects could have been produced by two or more contaminants, i.e., SO2 and its transformation products in combination with other PM components (Schrenk, 1949; Hemeon, 1955).

The London smog episode of 1952 resulted in an estimated 4000 excess deaths. Hospital admissions increased dramatically, mainly among the elderly and those with preexisting cardiac or respiratory disease. Otherwise healthy pedestrians, their vision limited to as little as three feet, covered their noses and mouths in an attempt to minimize their exposure to "choking air" (United Kingdom Ministry of Health, 1954). As a consequence of the 1952 London episode, daily measurements of British Smoke (BS), related to PM and SO2, started in 1954. These historical pollution data indicate that extremely elevated daily acidic aerosol concentrations (approximately 400 µg/m3) may be associated with excess human mortality when present as copollutants with elevated concentrations of PM and SO2. At non-episode pollution levels (H2SO4 < 30 µg/m3), associations between acidic aerosols and mortality in London are statistically significant even though these associations cannot be separated from BS or SO2. Increased hospital admissions for respiratory causes were also reported (Ito, 1993). Studies in the northeastern United States and Canada, where high levels of acidic aerosols are present during the summer, indicate that the increase in respiratory hospital admissions associated with acidic PM10 is about six times that for nonacidic PM10 (Thurston, 1994).

Short-term exposures to sulfuric acid (H2SO4) aerosols (0.5 µm in diameter) at ambient levels can alter mucociliary clearance, the primary lung defense mechanism. In healthy and asthmatic adults, mucociliary clearance is initially increased at 100 µg/m3 of H2SO4 and then decreased for higher concentrations (300-1,000 µg/m3 ) (Leikauf, 1984; Spektor, 1985). Pulmonary function was also decreased in adult asthmatics at those exposures (Spektor, 1985). Animal studies support these findings and also show altered resistance to bacterial infection and altered alveolar macrophage function. Low-level H2SO4 (100 µg/m3) reacts synergistically with O3 by exacerbating the O3 lung-function effects. In these controlled laboratory studies, the H2SO4 exposures have been controlled nasal exposures (normal daily human exposure is also by mouth) where a significant acid neutralization by ammonia (NH3) may occur, thus reducing the deposited lung dose (Schlesinger, 1992a, 1992b).


There are no reported toxicological studies of acute effects of inhaled metals at or below 0.5 µg/m3 . Most of the data originates in occupational settings and laboratory animals. These have limitations, such as how to extrapolate from animal models to humans; how to separate effects from the confounders probably present in occupational settings such as toxic gases and other inhalable particulate matter; and how to extrapolate effects from high- to low-exposure levels. Still, a review of the available literature regarding the health effects of trace elements at the lowest concentrations is included here. The mean concentrations of trace elements in the PM10 distribution--measured at Camp Thunderock, the site with the highest measured levels in the Gulf War region--still are too low to yield any known health effect. These measurements are shown in Table 3.5. For illustration purposes, the range of values in urban, rural, and remote areas of the United States is included. The Camp Thunderock data, except for nickel, fall in the range of rural areas, while nickel is within the urban range. It should be stressed that the levels in the United States are not high, but that levels in the Persian Gulf were very low.

Table 3.5
Concentration of Elements Associated with PM in the Ambient Air

United States
Elements Urban Rural Remote Camp Thunderock (August 1991)
Arsenic 2-2,300 1-28 0.007-1.9 4.25
Cadmium 0.2-7,000 0.4-1000 0.003-1.1 4.30
Chromium III 2.2-124 1.1-44 0.005-11.2 44.0
Iron 130-13,800 55-14,530 0.62-4160 8390
Nickel 1-328 0.6-78 0.01-60 136.0b
Lead 30-96,270 2-1700 0.007-64 587.0
Vanadium 0.4-1460 2.7-97 0.001-14 38.8
Zinc 15-8328 11-403 0.03-460 107.0


NOTE: Table 3.5 was updated November 21, 2005, to correct an error in the original table for the element Lead. In the original table, the concentration for Urban and Remote were transposed: the concentration for Urban was incorrectly shown in the Remote column and the number for Urban was in the Remote column. The numbers shown in red are the correct numbers. We have maintained the original Table 3.5 for historical purposes.


ang/m3 = 0.001 µg/m33.

bMean concentrations in the Gulf region were in the range of rural areas in the U.S. except for nickel, which was at the urban level.


Most of the available human inhalation data is based on occupational exposures to arsenic trioxide. For exposure concentrations above 1,000 µg/m3, symptoms include severe irritation of the nasal mucosa, larynx, and bronchia (Holmqvist, 1951; Pinto, 1953). These irritations may lead to hoarseness, laryngitis, bronchitis, and sometimes perforation of the nasal septa (Pinto, 1953). Increased peripheral vasospastic and Raynaud's syndrome were found in Swedish arsenic workers (Lagerkvist, 1986).


As the production of missiles, nuclear devices, and electronics grew and modern industrial technologies emerged, the risk of occupational exposure to beryllium became widespread. The environmental burden also increased as a result of emissions from plants producing and processing beryllium or its alloys and compounds. The major exposure to beryllium is through inhalation, which induces specific sensitization and nonspecific effects leading to chronic beryllium disease (CBD). CBD is an immunologically mediated granulomatous and fibrotic pulmonary disorder with increased numbers of lymphocytes in bronchoalveolar lavage fluid similar to that found in hypersensitive pneumoconitis. It has been hypothesized that epithelial injury and permeability changes occur early in CBD and are indicative of disease severity (Inoue, 1997). Associated symptoms are dyspnea on exertion, cough, chest pain, weight loss, and general weakness.


Health effects of cadmium exposure at ambient airborne concentrations have not been reported. The observed health effects in humans and animals are at concentrations three or more orders of magnitude higher than ambient. Above a critical threshold of 1000 µg/m3 per year, there is evidence of kidney damage, i.e., proteinuria (Mason, 1988). Acute respiratory effects of inhaled cadmium have been reported as pneumonitis and edema for exposures of approximately 300 µg/m3 for extended periods. Several studies found a statistically significant excess risk of lung cancer in the highest exposure groups (Elinder, 1985; Sorahan, 1987; Thun, 1985); thus, the International Agency for Research on Cancer (IARC) classified cadmium as a human carcinogen (IARC, 1993). Similar results were obtained from animal studies confirming that inhalation exposure to cadmium compounds can result in respiratory tract injury. Exposure to very high levels of cadmium has been shown to lead to a decreased immune response in mice (Graham, 1978).


Chromium, like many transition metal elements, is essential to life at low concentrations, yet toxic at higher concentrations. In addition to the overt symptoms of acute chromium toxicity, delayed manifestations of chromium exposure become apparent by subsequent increases in the incidence of various human cancers. Studies show conclusively that chromium in its hexavalent form (Cr (VI)) is both toxic and carcinogenic; in its trivalent form (Cr (III)) it is not, and moreover is essential in the metabolism of insulin (Bencko, 1985). Chromium (VI) easily crosses cell membranes and exerts genotoxic effects. The available evidence strongly indicates that Cr (VI) reduction in body fluids and long-lived non-target cells is expected to greatly attenuate its potential toxicity and genotoxicity, to imprint a threshold character to the carcinogenesis process, and to restrict the possible targets of activity.

On the other hand, no metabolic oxidation of Cr (III) has been observed. The Cr (VI) sequestering capacity of whole blood and the reducing capacity of red cells explain why this metal is not a systemic toxicant except at very high doses. Reduction by fluids in the digestive tract, i.e., saliva and gastric juice, and sequestering by intestinal bacteria account for the poor intestinal absorption of Cr (VI). Chromium (VI) escaping reduction will be detoxified in the blood and liver. These processes explain the poor toxicity of Cr (VI) and its lack of carcinogenicity when introduced orally or swallowed following reflux from the respiratory tract. The chemical environment in the gastrointestinal tract and the blood is effective even under fasting conditions in reducing Cr (VI) to one or more forms of Cr (III) (Kerger 1997). Inhaled Cr (VI) is reduced in the epithelial lining fluid and in the pulmonary alveolar macrophages. The lung parenchyma has reducing capacity with slightly higher specific activity than the bronchial tree. Therefore, even in the respiratory tract, the only consistent target of Cr (VI) carcinogenicity has barriers hampering its carcinogenicity. This protection could be overcome only by massive exposure through inhalation, as in work environments lacking proper industrial hygiene (De Flora, 1997).


Most of the human data are based on occupational exposures to iron oxide, with effects limited to respiratory symptoms and dysfunction, and no data are available on acute exposures. Mining and smelting processes generate iron oxides, silica, and other substances. Iron oxides deposited in the lung result in changes in lung X-rays. This effect is known by many names: siderosis, iron pneumoconiosis, hematite pneumoconiosis, iron pigmentation of the lung, and arc welder's lung. Siderosis is prevalent in 5-15 percent of iron workers exposed for more than five years (Sentz, 1969) and in a reported 34 percent in workers exposed to ferric oxide dust at concentrations ranging from 3500 to 269,000 µg/m33 (Teculescu, 1973). Evidence of lung fibrosis was not observed, but chronic cough was reported by 80 percent of the workers. Several studies reported high incidence of lung cancer mortality, but in all cases there was co-exposure to other potential carcinogens (Boyd, 1970). Animal studies that resulted in respiratory-tract cell injury and alveolar fibrosis were conducted at concentrations of 14,000 µg/m33 for one month (Nettesheim, 1975). Iron oxide particles have been the carrier particles for radioactive tracers, i.e., Technetium (99mTc) and Gold (198Au), in many human and laboratory animal studies designed to measure different aspects of pulmonary-particle deposition and clearance. These exposures were brief, but the concentrations were orders of magnitude higher than ambient levels. There are no reports of acute effects (Leikauf, 1984; Brain, 1991).


Children show a greater sensitivity to lead's effects than do adults, because children absorb and retain more lead in proportion to their weight. The most sensitive target of lead poisoning is the developing brain. Lead exposure at age two will result in continued deficits in neurologic development, such as lower IQ scores and cognitive deficits, at age five (Needleman, 1990; Schwartz, 1987). The primary sources of environmental exposure to lead are leaded paint, auto emissions, drinking water from plumbing leachate, and ceramic ware. Once in the bloodstream, lead is distributed in the blood, soft tissue, and mineralizing tissue. Bones and teeth of adults contain more than 95 percent of the body's lead content. Lead affects primarily the peripheral and central nervous system, blood cells, and the metabolism of vitamin D and calcium; it also causes reproductive toxicity (Gerber, 1980). Levels of lead in ambient air range from 7.6 x 10-5 µg/m3 in remote areas like Antarctica (Maenhaut, 1979) to more than 10 µg/m3 near sources such as a smelter, with an average annual concentration of below 1 µg/m3 for urban sites. The NAAQS for lead is 1.5 µg/m33 (EPA, 1997).


Mercury is ubiquitous in the environment, and it is also released in industrial activities and combustion of fossil fuels. Mercury so released is in an inorganic form, predominantly as metallic vapor. In aquatic environments, it is microbiologically transformed into the organic compound methylmercury. Therefore, populations with higher intake of foods originating in water have higher exposure to methylmercury (Hansen, 1997). In the past, methylmercury compounds were manufactured as fungicides or as unwanted byproducts in the chemical industry. Methylmercury absorbed from the diet distributes within a few days to all tissues in the body. It crosses the blood-brain and placenta barriers to reach its main target tissue, the brain. The biological half-life in human tissues is about 50 days, and mercury is eliminated chiefly in the feces after conversion to inorganic forms. Adult poisoning is characterized by focal damage to discrete anatomical areas of the brain such as the visual cortex and granule layer of the cerebellum. A latent period of weeks or months may lapse before signs and symptoms appear. The latter manifestation is paresthesia, ataxia, constriction of the visual fields, and hearing loss. Presently a chief concern is with the more subtle effects arising from prenatal exposure such as delayed development and cognitive changes in children (Clarkson, 1997).


In order of abundance in the earth's crust, nickel ranks as the 24th most common element, and it is found in different media in all parts of the biosphere. Nickel is a useful metal, particularly in various alloys, in batteries, in nickel plating, as a catalyst, and in pigments. Occupational exposure may lead to the retention of 100 µg per day. Environmental levels depend on emissions from nickel-manufacturing industries and airborne particles from combustion of fossil fuels. Absorption from ambient pollution is a minor concern. Vegetables, legumes, and nuts contain nickel, and the average dietary intake is 200 to 300 µg per day (Grandjean, 1984). Owing to the low absorption rate, nickel compounds in the gastrointestinal tract (except for nickel carbonyl) are essentially nontoxic after ingestion.

Some nickel compounds have been found to be carcinogens. Nickel carbonyl is correlated to nasal and lung cancer. Nickel subsulfide may be the most potent nickel carcinogen, but exposures to this compound are limited to specific occupations (Norseth, 1980). Other clinical manifestations include acute pneumonitis from inhaled nickel carbonyl, chronic rhinitis and sinusitis from inhaled nickel aerosols, and dermatitis and other hypersensitivity reactions from dermal exposures to nickel alloys (Sunderman, 1977). Cutaneous nickel allergy (contact dermatitis) affects 15-20 percent of the female general population and 1 percent of the males (Savolainen, 1996).


Most of the reported exposures are to vanadium pentoxide dusts in occupational settings. Worker exposure to vanadium dusts ranging in duration from several hours to several years, in concentrations from 100 to 300 µg/m3, produced temporary mild respiratory distress, i.e., productive cough, wheezing, chest pain, runny nose, or sore throat (Lewis, 1959). At concentrations ranging from 1,000 to 6,500 µg/m33 for 1-2 years, lower respiratory tract effects were observed, such as rhinitis, nasal discharge, irritated throat, bronchopneumonia, and asthmatic bronchitis (Sjoberg, 1950; Levy, 1984). Relatively low acute exposures--60 µg/m33 and 100 µg/m33 for 8 hours--produced cough and mucus formation that lasted for one week (Zenz, 1967). Acute and chronic laboratory animal studies show that the respiratory tract is the main target of inhaled vanadium compounds (Lee, 1986; Knecht, 1985).


Inhalation of zinc (mostly from zinc oxide fumes) may produce significant pulmonary irritation and inflammation, also known as metal fume fever. Since zinc has low toxicity, the exposure concentrations have to be in the mg/m3 range to induce these symptoms. Fever, chills, chest tightness, muscle/joint pain, sore throat, headache, and increased airway resistance were reported 4 to 8 hours after exposure to 4900 µg/m3 for 2 hours. Multiple higher-exposure concentrations resulted in the development of adaptation or tolerance after the initial symptoms of zinc fume fever subsided (Gordon, 1992).

Photochemical Pollution

Photochemical smog, or pollution, arises from a series of complex atmospheric reactions that result in a mixture of ozone, nitrogen oxides, aldehydes, peroxyacethyl nitrates, and reactive hydrocarbons. If sulfur dioxide is present, sulfuric acid droplets may be formed, as nitric-acid vapor can be formed from NO2. Hydrocarbons, under normal circumstances, are of much less concern because their concentrations are too small to become toxic. However, they are important because of the role they play in the formation of photochemical smog. In simple terms, ultraviolet light (UV) splits molecular oxygen, O2, into atomic oxygen, O, to combine with other O2 molecules to form O3. At the same time, in the troposphere, NO2 absorbs UV to form O and NO. This O combines with O2 to form O3 and O3 with NO forms NO2. This process is cyclic. In the absence of hydrocarbons, this series of reactions would reach a steady state with no excess of O3. The hydrocarbons are attacked by the free atomic O resulting in oxidized compounds and free radicals that react with NO to produce more NO2 and therefore a build-up of O3. This process occurs when the automobile emissions of morning commuters interact with the sun's UV to produce smog by noon.


Ozone (O3) is found in the free troposphere and in the planetary boundary layer (PBL), the layer next to the surface of the earth. Background O3 in the PBL occurs as the result of incursions from the stratosphere and through photochemical formation from precursors, CO, VOCs, and NOx. These precursors are all associated with combustion processes.

Short-term O3 exposure in healthy humans induces changes in pulmonary function, decreased volumes and flows, increased airway responsiveness, and airway irritation such as cough or pain on deep inspiration. Asthmatics experience similar effects and increased wheezing (Linn, 1994; Koenig, 1987, 1988; Molfino, 1991; Kulle, 1984). Inflammatory responses have been observed after acute exposures to O3 at concentrations found in U.S. cities (Devlin, 1990, 1991, 1996; Koren, 1990, 1991). Recovery from acute exposure is usually complete within 24 hours of the end of exposure. There is evidence of a plateau in lung volume in response to prolonged O3 exposure. Also, available data indicate that exposure to O3 for months and years causes structural changes in several regions of the respiratory tract. Research to date indicates that the area most affected is the centriacinar region, where alveoli and conducting airways meet (Sherwin, 1991).

There are very few human studies on binary pollutant exposure. Ozone in combination with SO2, H2SO4, HNO3, NO2, or peroxyacetyl nitrate (PAN) causes an additive response on lung spirometry or symptoms (Dreshsler-Parks, 1989; Aris, 1991; Hazuka, 1994; Utell, 1994; Koenig, 1990, 1994). Animal studies of O3 and NO2 or H2SO4 show that effects can be additive, synergistic, or even antagonistic, depending on the endpoint studies (Gelzleichter, 1992; Warren, 1986; Schlesinger, 1992a, 1992b). The chronic effects of co-pollutant exposure are still not understood. There is evidence suggesting that people with preexisting limitations in pulmonary function and exercise activity, e.g., asthma, COPD, chronic bronchitis, and ischemic heart disease, are at risk from O3 exposure.

Nitrogen Dioxide

Nitrogen dioxide (NO2), like O3, is a deep-lung irritant, but less potent. No alterations in lung function were observed in healthy humans after exposure to NO2 for up to 4 ppm for three hours; at 1.5-2 ppm, however, slightly enhanced airway reactivity was shown (Mohsenin, 1987). It is interesting to note that ascorbic-acid pretreatment protected the subjects from this hyperreactivity (Kjaergaard, 1996). Animal studies have shown associations between NO2 and both viral and bacterial infections, suppression of macrophage, and alterations in lung clearance (Schlesinger, 1987). NO2 is also an important indoor air pollutant, especially in homes with unvented gas stoves or kerosene heaters.

Sulfur Dioxide

Sulfur dioxide (SO2) is a gas formed when fuels containing sulfur, mainly coal and oil, are burned, and also during smelting and other industrial processes. SO2 readily oxidizes to sulfate in the atmosphere in the presence of catalysts--e.g., metals such as iron, manganese, and vanadium in dispersing smokestack plumes, or via photochemical processes. However, most of the oxidation of SO2 occurs in the atmosphere, where it is transformed into sulfuric acid (H2SO4). As such, it may undergo long-range transport to areas hundreds of miles away from the emission source. SO2 is an upper-airway irritant that can stimulate bronchoconstriction and mucus secretion. Animal studies indicate that relatively low concentration exposures of SO2 (0.1 to 20 ppm) for long periods have marked effects consistent with bronchitis (Nadel, 1965).

Carbon Monoxide

Carbon monoxide (CO) comes from both natural processes (approximately 40 percent, mostly through oxidation of hydrocarbons but also from plants and oceans) and from anthropogenic processes such as combustion of fossil fuels and oxidation of methane (approximately 60 percent). CO, an asphyxiant, is readily absorbed through the lungs into the blood stream, where it competes with O2 to bind to hemoglobin (Hb) in red blood cells, forming carboxyhemoglobin (COHb). Its toxicity arises from its high affinity to Hb, approximately 240 times greater than that of O2 (Haldane, 1898). A unique characteristic of CO exposure is that the blood COHb level is a very useful biomarker of exposure. The level of COHb may be determined directly by blood analysis and also estimated by measuring the CO concentration in exhaled breath. COHb levels can be calculated from known CO exposures by solving the nonlinear differential equation known as the Coburn equation (Coburn, 1965).

There is a baseline blood level of COHb of approximately 0.5 percent for healthy adults and 1-8 percent for smokers. The most sensitive members of the general population to CO exposure are those with ischemic heart disease. They will start experiencing reduced exercise duration due to increased chest pain (angina) at CO levels that will give them 3-6 percent COHb (Kleinman, 1989; Allred, 1991). Healthy individuals will experience a reduction in their maximal exercise performance at CO levels that, after one hour, give them COHb levels of 2-3 percent. At COHb blood levels greater than 5 percent, healthy individuals may have equivocal effects on visual perception, audition, motor and sensory performance, vigilance, and other measures of neurobehavioral performance. At higher blood COHb levels, 10 percent or more, neurological effects could occur, including headache, dizziness, weakness, nausea, and confusion. Unconsciousness and death can occur with continuous exposure to high CO levels in a workplace or in unventilated rooms with faulty unvented combustion appliances.

Hydrogen Sulfide

The acute effects of exposure to H2S are well recognized. Odor of "rotten eggs," followed by olfactory paralysis, mucosal irritation, and keratoconjunctivitis are the typical effects of H2S at lower concentrations. H2S-induced acute central toxicity leads to reversible unconsciousness called a "knockdown." It has been suggested that repeated or prolonged knockdowns are associated with chronic neurological sequelae (Guidotti, 1994). Knockdowns can be fatal due to respiratory paralysis, cellular anoxia, and pulmonary edema. There are some indications that other chronic health effects include neurotoxicity, cardiac arrhythmia, and chronic eye irritation, but apparently not cancer. Concentrations above 50 ppm--five times the occupational limit--can cause death (Kilburn, 1995). Healthy human volunteers inhaled air with 10 ppm H2S in a blind test in-between 30-minute exercise sessions at 50 percent of their maximal oxygen. All experienced a significant decrease in oxygen uptake with concomitant increase in blood lactate. However, no significant changes were observed in arterial blood parameters (Bhambhani, 1997), pulmonary function, or diffusion capacity (Bhambhani, 1996).

Long-Term Exposure to Air Pollution (Cancer)

The role of inhaled pollution in human lung cancer is difficult to assess because the vast majority of respiratory cancers result from cigarette smoking. Volatile organic compounds and nitrogen-containing and halogenated organic compounds account for most of the compounds that have been studied with animal and genetic bioassays. Most of these compounds are derived from combustion sources, from power plants to incinerator emissions. Other potential carcinogens also result from mobile sources as products of incomplete combustion and their atmospheric transformation products, as well as fugitive or accidental chemical releases.

The carcinogen potency of air pollution resides usually in the particulate fraction. Polycyclic organic chemicals and semivolatiles are associated with the particulate fraction and could have a prolonged residence time at sensitive sites in the respiratory tract when inhaled. Genetic bioassays have demonstrated potent mutagenicity, and presumably carcinogenic potential, of various chemical fractions of ambient aerosols. Copollutants such as irritant gases that initiate inflammation may promote carcinogenic activity (Lewtas, 1993).

Populations at Risk

Groups within the general population most clearly at risk include the elderly and those with cardiopulmonary diseases. Epidemiological studies indicate that mortality and morbidity from respiratory causes are strongly related to ambient PM exposures. Possible mechanisms include airway narrowing, increased mucous secretion or increased viscosity, and inflammation and epithelial cell damage in persons with respiratory disease. Cardiac arrhythmia has been hypothetically linked to mortality due to acute PM exposure. Risk of mortality due to lower-respiratory disease, e.g., pneumonia, is increased by ambient PM exposure by both exacerbation and increasing susceptibility to infectious disease by decreasing clearance, impairing macrophage function, or other specific or nonspecific effects on the immune system. Smokers comprise about 80 percent of individuals with COPD, and together with a small but notable portion of patients with cardiovascular disease, are likely to be at increased risk for PM health effects. Asthmatics are more responsive than non-asthmatics to acidic aerosols; asthma exacerbation, even requiring medical care, is associated with PM10 exposure.

Health Effects Studies Conducted During the Oil Fires

Only a few quantitative studies were performed during the Persian Gulf War to assess the possible health effects of the oil well fires. Four are described here. One compared U.S. personnel in Kuwait City, oil firefighters, and a control group. Another analyzed the genotoxicity of the soot. A third study examined feral cats in Kuwait City and Ahmadi. The fourth was comprised of several self-administered health questionnaires. The USAEHA also estimated cancer and noncancer risks due to the environmental exposure levels it measured using an EPA methodology[4], finding the risks similar to those for the general U.S. population (USAEHA, 1994).


This study documented VOC levels in the blood of U.S. personnel in Kuwait City in May 1991 (Group I) and American firefighters working in the oil fields in October 1991 (Group II). Concentrations of VOCs in both groups were compared with a random sample of persons in the United States (Control Group). The median concentration of VOCs in Group I were equal to or lower than those in the Control Group. However, significant differences were found with Group II. Median levels of ethyl-benzene were 10 times higher than in the Control Group, while benzene, toluene, and xylene levels were more than double those of the Control Group. These VOC levels are comparable to those measured in German workers in Kuwait. The results were drawn from a small sample of volunteers. Considering that the half-life of these compounds in blood is less than 4 hours, they may be taken only as a suggestion of what the VOC-exposure levels were in Kuwait during the fires (Etzell, 1994). This work is summarized in Table 3.6.

Table 3.6
VOC Concentrations in Blood in U.S. Personnel

VOC Kuwait City Personnel
(Group I)
(Group II)
U.S. Reference
Benzene 0.035 0.18 0.066
Ethyl-benzene 0.075 0.53 0.052
m,p-Xylene 0.14 0.41 0.18
o-Xylene 0.096 0.26 0.10
Toluene 0.24 1.5 0.30

SOURCE: Etzell, 1994.


The in-vitro genotoxicity of soot from the Kuwait oil fires was also studied. Dose-dependent increases were observed for both sister chromatoid exchanges in human peripheral blood lymphocytes and mutation of the hprt locus in the metabolically competent human lymphoblast cell line AHH-1. Similar magnitudes were observed when testing air particulate isolated from Washington, DC. Using 32p-postlabeling assay, no increase in DNA adduct formation was observed in AHH-1 cells treated with particulates isolated from sampling in Kuwait (Kelsey, 1994).

Feral Cats

A study to assess histopathological lesions and analyze chemical accumulations from exposure to the smoke from the Kuwait oil fires collected 12 adult feral cats from Kuwait City and 14 from Ahmadi. Specimens from the lungs, liver, and kidneys, as well as urine and blood samples were examined. The pharyngeal mucosa in all animals was normal. Neither hyperplasia of the epithelium nor cellular atypia were observed. Minimal changes were observed in the larynx. Only two of the 26 cats examined had hyperplasia of the submucosal glands and both were from Kuwait City. No major organs displayed lesions that could be attributed to breathing smoke from the oil well fires or to hydrocarbon inhalation.

Although the results from the heavy-metal analysis were difficult to interpret, because of a lack of "normal" ranges for cats as well as ignorance about what is "normal" for the region, the results seem reasonable when compared to other animal species. Analyses for vanadium and nickel were within normal limits, suggesting that the smoke was not a serious health risk. The clearance mechanism of the ciliated region of the lung apparently was efficient at eliminating particulate matter from the smoke. None of the lesions associated with prolonged exposure to respiratory irritants, and consistent with chronic bronchitis, chronic obstructive pulmonary disease, pulmonary fibrosis, or emphysema, was observed (Moeller, 1994).

Health Survey Questionnaires

A self-administered symptoms and medical-history questionnaire was given to the 11th Armored Cavalry Regiment, then based in Fulda, Germany. A group of 331 soldiers (Group I) completed a survey prior to deployment during the period beginning just before departure from Germany and ending about eight weeks after arrival in Kuwait, a total of 12 weeks. Another 1599 soldiers (Group II) completed a postreturn survey about four weeks after returning to Germany, which included questions on symptoms before, during, and after the mission. The symptoms that appeared once or a few times during the first eight weeks in Kuwait but were not experienced before going to Kuwait were headache (55 percent), lightheadedness (48 percent), fatigue or weakness (45 percent), skin rashes (41 percent), and diarrhea (42 percent). About 35 percent of those who did not usually have cough or phlegm first thing in the morning before deployment reported the symptom after arriving in Kuwait.

The respondents in Group II recalled their complaints in Kuwait and compared them to those following their return. They reported that the persistence of symptoms was higher in Kuwait than during the eight weeks following their return to Germany. The increase in complaints (percent in Kuwait minus percent in Germany) for occasional symptoms were: eye irritation (27 percent), burning eyes (25 percent), shortness of breath (17 percent), weakness or fatigue (15 percent), skin rashes (14 percent), and respiratory irritation (14 percent). Symptoms were related to reported proximity to the oil fires, and their incidence generally decreased after the soldiers left Kuwait (Petruccelli, 1997).

Another self-administered symptoms questionnaire was developed by the U.S. Navy, and was completed by 2668 servicemen during March 28-31, 1991. Females were excluded because of their very small number. The respondents were divided into three categories according to their proximity to the oil well fires and duration of exposure. Group I consisted of 892 Marines who had the longest exposure, about five weeks at the time of the survey, and were closest to the burning oil wells. Group II included 978 Marines who had a short exposure to the oil fires, and were then stationed in Saudi Arabia about 120 km south of the Kuwait border. Depending on wind conditions, smoke from the oil fires would still have been clearly visible. Group III consisted of 831 Marines who had no direct exposure to the oil fires, having been located in southern Kuwait about 200 km south of the nearest oil field, where only a distant haze was visible on the horizon.

Marines in Group I reported the highest rate of gastrointestinal episodes and respiratory symptoms, followed by Group II. Similar patterns were observed for burning and red eyes. Adjusting for flu vaccination, history of respiratory disease, and smoking status, Group I reported wheezing, coughing, runny nose, and sore throat significantly more frequently than Group III. Group II reported significantly fewer colds than Group III. Smokers reported more complaints than nonsmokers. No patterns were observed when prevalence of respiratory symptoms within groups were examined by job class. The prevalence of wheezing, coughing, and runny nose for each group decreased from Group I to Group II to Group III, the latter being away from the smoke and on the coast away from blowing sand and dust (DoD, 1993).

A study was undertaken to assess and monitor the effects of the oil well fires on a squadron of 125 British bomb-disposal engineers who remained in Kuwait for five months. All subjects completed a health questionnaire and had their respiratory function measured. From June to October 1991, measurements were taken every two weeks; the subjects repeated the questionnaire just before returning to Britain. When data were stratified according to either smoking history prior to deployment or by amount smoked during the tour, no significant differences in either lung-function indices or symptoms were detected when compared to predeployment values (Coombe and Drysdale, 1993).

Studies Conducted After the Gulf War

CCEP Study

The Department of Defense Comprehensive Clinical Evaluation Program (CCEP) provides medical evaluation to Gulf War veterans now on active duty or retired, members of the reserve components who are Gulf War veterans, and eligible members of families who are experiencing illnesses that they allege to be related to their service in the Persian Gulf. As of December 1995, more than 27,000 individuals had enrolled in the program, and as of April 1996 18,598 had completed the evaluation process and had their health records entered into the CCEP database. CCEP participants report a wide variety of symptoms spanning multiple-organ systems in no consistent, clinically apparent pattern. The symptoms being reported in the CCEP are not unique; fatigue, joint pain, headache, or sleep disturbances are common among CCEP participants. The frequency distribution by category of diagnosis assigned by the CCEP is presented in Table 3.7.

Table 3.7
Frequency Distribution of Diagnoses

Diagnostic Categories (ICD-9-CM Code) Primary Diagnosis:Malesc Primary Diagnosis:Femalesd Primary Diagnosis:Alle Any Diagnosis:Alle
Psychological conditions (230-219) 18.3 19.1 18.4 36.0
Symptoms, signs, and ill-defined conditions (780-799)a 18.1 16.5 17.9 43.1
Musculoskeletal system disease (710-739) 18.6 15.9 18.3 47.2
Healthyb(V65.5) 9.9 8.6 9.7 10.2
Respiratory system diseases (460-519) 6.9 6.1 6.8 17.5
Digestive system diseases(520-579) 6.5 4.9 6.3 20.4
Skin and subcutaneous tissue diseases (680-709) 6.3 6.0 6.2 19.9
Nervous system diseases (320-389) 5.3 8.8 5.7 17.8

aIncludes conditions categorized according to ICD-9-CM nomenclature of cases for which no diagnosis was classifiable elsewhere; no more specific diagnosis could be made; signs or symptoms proved to be transient; cases in which a precise diagnosis was not available.
bIncludes registered participants without complaint or sickness as well as those diagnosed as normal or healthy.

The order of these diagnoses was determined by usual clinical practice, basing the ranking on the most severe conditions relative to the patient's chief complaints. The most prevalent primary diagnostic categories, accounting for 67.7 percent of the participants, were psychological conditions (18.4 percent); musculoskeletal and connective tissue diseases (18.3 percent); symptoms, signs, and ill-defined conditions (17.9 percent); respiratory diseases (6.8 percent); and digestive system diseases (6.3 percent). Nearly 10 percent received a diagnosis of healthy. When both primary and secondary diagnoses were considered, similar patterns emerged. The most common categories were musculoskeletal diseases (47.2 percent); symptoms, signs, and ill-defined conditions (43.1 percent); psychological conditions (36.0 percent); digestive system diseases (20.4 percent); skin and subcutaneous diseases (19.9 percent); respiratory diseases (17.5 percent); and nervous system diseases (17.8 percent).

Up to seven diagnoses, including healthy, could be reported (one primary and up to six secondary). Among the participants, 19.9 percent had only 1 diagnosis, 18.7 percent had 2, the median was 3, the mean was 3.4, and 9.1 percent were given 7 diagnoses.

Birth Defects

A recently published study on the risk of birth defects among children of Persian Gulf War veterans analyzed the live-birth records at 135 military hospitals over three years (1991-1993). In this period, 33,998 infants were born to Gulf War veterans and 41,463 to nondeployed veterans in military hospitals. The overall risk for birth defects was 7.45 percent, and the risk for severe birth defects was 1.85 percent. There were no significant differences between children born to deployed and nondeployed veterans. When the same definitions of birth defects were applied to the approximately 320,000 live births that occurred nationwide in 1992, the risk of any birth defect was 8.4 percent and the risk of a severe defect was 1.9 percent. The birth defect rates for veterans parallel those of the general population. There was no significant association for either men or women between service in the Gulf War and risk of any birth defect (Cowan et al., 1997).

Iowa Study

Several studies have been conducted to assess prevalence of self-reported symptoms, illnesses, and psychiatric conditions among military personnel deployed in the Persian Gulf. The Iowa Persian Gulf Study surveyed 4886 subjects belonging to one of the following groups: Gulf War regular military, Gulf War National Guard/Reserve, non-Gulf War regular military, and non-Gulf War National Guard/Reserve. The results indicate that, compared to non-Gulf War military, Gulf War veterans reported a significantly higher presence of symptoms of depression, post-traumatic stress disorder (PTSD), chronic fatigue, cognitive dysfunction, bronchitis, asthma, fibromyalgia, alcohol abuse, anxiety, sexual discomfort, and diminished mental and physical functional health. Within the National Guard/Reserve, the differences in prevalence for the above-listed symptoms between Gulf War veterans and non-Gulf War veterans are higher than for comparable groups in the regular military. For asthma and bronchitis there is a statistically significant (2.3 percent) difference between Gulf War military and non-Gulf War military (Iowa Persian Gulf Study Group, 1997).

The Iowa Persian Gulf Study Group is performing a follow-up study that includes a complete medical evaluation of a small random sample of the original cohort to verify the sensitivity and specificity of the larger study. The medical evaluation includes pulmonary-function tests that will also aid in verifying the small difference in lung-disease prevalence between the two military groups (Iowa Persian Gulf Study Group, 1997).

One must look cautiously at these studies because none includes quantitative measures of exposure or effects; findings are based solely on self-reported survey questionnaires administered years after the war's end. Some survey studies conducted years later probably have large uncertainties due to different recall and other biases.


The military personnel who were deployed in the Gulf were healthy and, for the most part, young. Thus, one could argue that the preceding discussion on the possible health effects due to exposures to pollutants present in the Gulf would not apply to Gulf War veterans to the same extent they would to susceptible populations. Moreover, these veterans' exposure levels, except for PM10, were much lower than the U.S. occupational standards, and even lower than ambient air standards. The exposure concentrations of the pollutants measured in the Gulf, except for PM10, were lower than those in U.S. urban areas. Thus, one would not expect this population to experience ill health effects as a result of exposures in the Gulf.

Several factors may have made Gulf War veterans more vulnerable to pollutants. A significant number of Gulf War veterans were smokers and cannot be considered as having no lung impairment. Smokers react differently than nonsmokers to inhaled pollutants, even at a low concentration of irritants. Also, a small percentage of the veterans may have had some underlying predisposition to pulmonary disease such as asthma that might be triggered after exposure to high levels of PM10.

The levels of PM10 were extremely high, not because of emissions from the oil fires, but from the desert sand. These high levels of PM10 may perhaps explain the preliminary findings of the Iowa Study regarding respiratory symptoms as well as some of the respiratory symptoms reported in the CCEP. Although there are personal communications indicating that there were increased respiratory complaints among the indigenous population during the oil fires, no evidence of health effects or epidemiological studies were found in the peer-reviewed literature.

[1]Personal exposure is the concentration inhaled by a subject. It can be measured directly in the air around the mouth-nose of a subject, e.g., with a sampler attached to the lapel.

[2]Genotoxic: denotes a substance that may cause mutation or cancer by damaging DNA.

[3]Relative Risk (RR) is the ratio of probabilities that a disease will occur among those exposed to a factor to that of those not exposed.

[4]As part of the Superfund Program, the EPA developed reference dose values (RfCs) for many toxic agents that estimate a daily exposure level for the general population that is likely to be without appreciable risk during a specific period. These estimates have been used to calculate cancer and noncancer risks (EPA, 1992).

Previous Chapter
Next Chapter