Toxicological Review of Barium and Compounds

following administration of barium in drinking water at the highest level tested (150 mg/kg-day). 

Chronic and subchronic drinking water studies in rats and mice  (NTP, 1994; McCauley 

et al., 1985) provide evidence that the kidney is a sensitive target of barium toxicity.  NTP 

(1994) observed chemical-related nephropathy in mice following chronic or subchronic drinking 

water exposure to barium.  The lesions were characterized by tubule dilatation, renal tubule 

atrophy, tubule cell regeneration, hyaline cast formation, multifocal interstitial fibrosis, and the 

presence of crystals, primarily in the lumen of the renal tubules.  These changes were 

characterized as morphologically distinct from the spontaneous degenerative renal lesions 

commonly observed in aging mice (NTP, 1994).  Similar lesions were also observed in rats 

following subchronic exposure.  In the chronic rat study, spontaneous nephropathy was observed 

in the majority of animals in both control and treatment groups precluding the detection of any 

treatment-related effect.  Increased kidney weights were observed in male and female rats and 

female mice following 13 weeks of exposure.  Female rats were the only animals with increased 

kidney weights following 15 months of exposure.  

Several case reports (Seaton et al., 1986; Pendergrass and Greening, 1953) and a 

prospective study conducted by Doig (1976) have reported baritosis in workers exposed to 

airborne barite ore or barium sulfate.  Baritosis is considered a benign pneumoconiosis 

characterized by intense radiopacity of discrete opacities usually profusely disseminated 

throughout the lung.  Spirometric lung function tests were normal in the workers examined by 

Doig (1976).  Upon exposure termination, there was an apparent decrease in barium levels in the 

lung (Doig, 1976); the barium-related lesions are also potentially reversible (ACGIH, 1992). 

NIOSH (1982) reported an increased incidence of hypertension in workers exposed to an 

unspecified concentration of barium; these results should be interpreted cautiously because it is 

likely that the workers were also exposed to other metals, including lead, which has a known 

hypertensive effect. 

Data on the toxicity of inhaled barium to animals are limited.  Tarasenko et al. (1977) 

reported perivascular and peribronchial sclerosis with collagenation in the lungs and increases in 

arterial pressure in rats exposed to barium carbonate.  The deficient reporting of the methods and 

results (in particular, the lack of information on the aerosol generation, number of animals tested, 

incidence data, and statistical analysis) limits the usefulness of this study for hazard assessment. 

50

---

A reproductive toxicity study did not find a significant dose-response in gestation length, 

pup survival, or occurrence of external abnormalities in rats and mice exposed to barium 

chloride in drinking water (Dietz et al., 1992).  Based on the limited amount of data available, it 

is not possible to make a definitive conclusion about the potential for barium to impair 

reproductive functions. 

An area of scientific uncertainty concerning the noncancer hazard assessment for barium 

is identification of the most sensitive endpoint of barium toxicity in humans.  The results of the 

NTP (1994) drinking water studies in mice and rats suggest that renal toxicity is the most 

sensitive endpoint.  However, it is not known if a similar relationship would exist following 

chronic exposure in humans.  Another area of scientific uncertainty is whether any toxicological 

or toxicokinetic differences exist between children and adults.  Animal data (Cuddihy and 

Griffith, 1972; Taylor et al., 1962) suggest that gastrointestinal absorption may be greater in 

children than in adults. 

No oral human carcinogenicity data are available.  Oral exposure studies in rats and mice 

(NTP, 1994; McCauley et al., 1985; Schroeder and Mitchener, 1975a, b) did not find significant 

increases in tumor incidence following chronic exposure to barium. 

No inhalation carcinogenicity data are available for humans.  The inhalation and 

intratracheal studies in animals conducted by Tarasenko et al. (1977) are inadequate for 

carcinogenicity evaluation because of several deficiencies in the design and reporting, including 

single or subchronic exposure duration, inadequate reporting of aerosol generation methodology, 

deficient reporting of study results (including the apparent lack of statistical analysis), and the 

use of only one sex (males). 

Based on the weight of evidence, barium can be classified as Group D, not classifiable as 

to human carcinogenicity, using the 1986 guidelines (U.S. EPA, 1986c).  Although adequate 

chronic oral exposure studies in rats and mice have not demonstrated carcinogenic effects, the 

lack of adequate inhalation studies precludes assessing the carcinogenic potential of inhaled 

barium.  According to the proposed guidelines, barium would be considered not likely to be 

carcinogenic to humans following oral exposure, and its carcinogenic potential cannot be 

determined following inhalation.  The lack of adequate inhalation carcinogenicity data is an area 

of scientific uncertainty for this assessment. 

51

---

6.2.  DOSE-RESPONSE ASSESSMENT 

The chronic oral RfD of barium that is considered to be without deleterious noncancer 

effects is 0.2 mg/kg-day.  This value is based on an increased incidence of chemical-related 

nephropathy in male mice chronically exposed to barium chloride in their drinking water (NTP, 

1994).  The RfD was calculated by dividing the lower 95% confidence limit for the dose 

estimated to affect 5% of the population (BMDL

05

) by an uncertainty factor of 300.  The 

combined uncertainty factor of 300 accounts for uncertainty associated with extrapolation from 

laboratory animals to humans, variation in susceptibility among humans, and uncertainty 

resulting from limitations in the data base. 

The overall confidence in this RfD is medium.  Medium confidence in the RfD reflects 

the high confidence in the principal study but medium confidence in the data base. Confidence in 

the principal study is high because it is a high quality study conducted by the National 

Toxicology Program (NTP, 1994).  The study included a control group and three exposure 

groups, and each group contained 60 animals of both sexes.  Standard NTP quality assurance and 

quality control procedures, including a review of all histology data by the Pathology Working 

Group, were employed.  Confidence in the data base is medium because it lacks human data that 

define an adverse effect level but contains adequate dose response information for chronic and 

subchronic animal studies conducted in more than one species. 

At the present time, no adequate data are available to derive an RfC for barium.  The 

available human and animal data suggest that the respiratory tract may be a sensitive target of 

toxicity; thus, it would not be appropriate to derive an RfC for barium, based on oral data. 

Dose-response assessment for carcinogenic effects is not applicable because the oral data 

suggest that barium is not likely to be carcinogenic and the inhalation data are inadequate. 

52

---