Hydrogen sulphide

Endpoint summary

Description of key information

H2S is a colourless, extremely flammable gas, which is heavier than air. 
H2S exposure levels that result in temporary unconsciousness (e.g., 15-30 minutes) can cause profound neurophysiological, neurobehavioral, neurocognitive, neurophysical, respiratory, and opthalmologic deficits that are persistent. 
According to Daunderer: Klinische Toxikologie, 33. Erg.-Lfg. 1/88 small amounts of H2S are relatively harmless for the body, as they are rapidly oxidized in the human metabolism to sulfate and thiosulfate. High concentrations of H2S can lead to an immediate arrest of respiration. Lesser amounts can cause a pulmonary edema after 3 to 4 days.
The acute toxic mechanism  of H2S seems to be an inactivation of the cytochrome-oxydase by binding to the trivalent iron in the enzyme.  The intravenous injection of 250 mg 4-(Dimethylamino)pyridine (4-DMAP) is the effective antidot treatment.
Since this treatment was established no further fatal intoxications with H2S have been reported. 
By this antidot treatment also the delayed toxicity and the extent of late sequilae were markedly reduced.
The actual occupational exposure level is 5 ppm (7 mg/m³).

Additional information

As summarized by US EPA (2003).

Case Studies

A 19-year-old oil rig worker was exposed to unspecified concentrations of H2S (Burnett et al., 1977) rendering him unconscious for an indeterminate amount of time. Upon resuscitation, he exhibited malaise, anterior chest pain, dyspnea, headache, nausea and vomiting, tearing of the eyes and photophobia and coughed up blood. Upon arrival at the hospital for further treatment, his vital signs were normal and he was no longer in respiratory distress. He had severe photophobia and blepharospasms, but no signs of conjunctivitis. He also possessed a cough and some motor weakness of his right arm and leg. A neurologic examination and a chest x-ray revealed no abnormalities. After a 3-day stay in the hospital, he was discharged.

Two fatalities were due to massive aspiration of liquid manure and one fatality was due to severe pulmonary edema with no aspiration of the manure (Osbern and Crapo, 1981). Another 41-year-old individual fell unconscious into liquid manure during a rescue attempt. After resuscitation, he had difficulty breathing and was agitated, but he exhibited no focal neurologic deficits. His initial chest radiograph showed a five-lobed alveolar infiltrate. After two weeks in the hospital, his chest radiograph showed improvement and his lung function was normal except for a slightly reduced maximum midexpiratory flow rate. Six months after the accident, he had a normal chest radiograph and was asymptomatic. Blood sulfide levels of the two individuals who had fatally aspirated manure were 5.0 and 3.6 mg/L. The individual who had died by pulmonary edema had a blood sulfide level of 0.8 mg/L. Control blood samples exhibited sulfide levels below 0.05 mg/L while blood samples from random autopsy cases of badly decomposed subjects did not exceed 0.4 mg/L. No blood sulfide measurements were performed on the 41-year-old survivor. Eight days following the accident, air analyzed from the liquid manure storage tank contained 6,360 ppm methane, 400 ppm carbon monoxide, 1.5 ppm ammonia, 2% carbon dioxide, 18% oxygen, and 76 ppm (106 mg/m3) H2S.

Another case study illustrates the possible long-term sequelae of H2S exposure. A 30-year-old man displayed dyspnea, chest tightness, and haemoptysis following exposure to a toxic gas in a lavatory facility at his place of work (Parra et al., 1991). The facility was connected to a manure pit; no measurements of H2S were performed. Physical examination and routine laboratory studies revealed no abnormalities. However, a chest radiograph detected a mild bilateral interstitial pattern. Bronchoscopy showed a widespread reddish mucosa. Pulmonary function tests showed a mild restrictive disease. After five months, the patient possessed residual exertion-dyspnea but was otherwise asymptomatic. The diagnosis was pneumonitis caused by the inhalation of a toxic gas. Other exposed workers exhibited nausea, vomiting, dizziness, dyspnea, and eye and nose irritation. One of the exposed workers died a few hours after exposure. An autopsy revealed haemorrhagic bronchitis and the cause of death was asphyxia attributed to the inhalation of a “toxic gas,” with H2S suspected as a major component.

Several case studies report on the rapid toxicity following exposure to high levels of H2S. A 14-year-old boy found a discarded cylinder containing H2S in a rubbish dump and immediately died when he opened the tank (Allyn, 1931). His father also died during a rescue attempt. Both bodies were deeply cyanosed. A fatality, apparently due to release of H2S from a load of refinery waste was described by Simson and Simpson (1971). This report also mentioned eyewitness accounts of intense central cyanosis in the victim.

At a poultry feather fertilizer plant, a worker was exposed to H2S while attempting to repair a leak and was killed (Breysse, 1961). Hydrogen sulfide was created as a byproduct of the putrefaction of the feathers, and was eliminated through a pipe leading to a sawmill log pond where it was discarded. In the lungs of the victim, the alveolar spaces were filled with edema fluid and numerous pigment-filled macrophages. The diagnosis was pulmonary edema, and the cause of death was H2S inhalation. Measurements of H2S concentrations at various locations in the fertilizer plant revealed levels as high as 4,000 ppm (5,560 mg/m3) during the cooking and putrefaction of feathers.

A 16-year-old boy suffered fatal H2S exposure during transport of liquid manure (Hagley and South, 1983). He was found at the bottom of the manure tank, pale, unconscious and apneic. There was no evidence that he had aspirated any manure. The boy began breathing following resuscitation and his color returned to normal. His heart rate and blood pressure were normal though he did not respond to painful stimuli. Over the following hour, he became responsive to stimuli, but then developed extensor spasms and began to hyperventilate. A chest radiograph and a CT scan of the head were normal. The patient developed pneumonia. However, his neurological condition deteriorated, and he suffered partial seizures and exhibited a decerebrate response to painful stimuli. The patient died five days after the accident with clinical signs of brain stem damage. Postmortem examination revealed that he had cerebral edema. A week later, H2S was measured 30 cm below the manhole of the tank and was found to exceed the upper limit of detection of the monitoring equipment at 150 ppm (208 mg/m3).

By contrast, Milby (1962) has reported two cases of severe, but non-fatal H2S poisonings. While disposing of a H2S-filled gas cylinder and the subsequent release of its contents, two men that were exposed to the gas collapsed immediately and underwent convulsions while unconscious. Upon being rescued from the gas, the men required artificial respiration to begin breathing on their own. Both men were hospitalized in an unconscious state and were later revived. At the conclusion of their hospital stay, both men were without symptoms. The author states that they “remained well to the present,” but it is unclear how long an interval that represents.

A mass exposure to H2S took place among workers laying the foundation of a municipal sewage pumping station, leading to the death of a police officer attempting to rescue an unconscious worker and persistent neurological sequelae in at least one other individual (Snyder et al., 1995). Complications included: decreased ability to communicate, slow speech, marked visual memory impairment with poor acquisition, and difficulty retaining and recalling new information. The findings were reported as essentially unchanged 12-18 months after exposure. Physical findings in affected individuals were limited to pharyngeal and conjunctival erythema and corneal abrasions. Additional reports associated H2S exposure (via sewer gas) with lung function and neurological decrements and, in certain studies, deaths were reported by Watt et al. (1997) and Hall and Rumack (1997).

Tvedt et al. (1991a) described the follow-up of six H2S-exposed patients who lost consciousness and experienced signs of long-term neurological sequelae associated with hypoxic brain damage. Immediate symptoms included cyanosis, pulmonary edema, seizures, and coma. Delayed symptoms (5-10 year follow-up) ranged from worker disability due to neurological symptoms to brain damage severe enough to qualify as dementia in one individual. Five patients who had been unconscious in an H2S atmosphere for 5-20 minutes showed persistent impairment with memory and motor function most affected. The two patients with the most serious symptoms developed pulmonary edema. No measures of H2S in air, blood or urine were reported. The investigators concluded that the outcome of acute exposures can be variable and that neurological symptoms may not become apparent until a considerable period of time has elapsed.

Tvedt et al. (1991b) also described delayed neuropsychiatric symptoms in an individual who became unconscious for 15-20 minutes during an H2S exposure. He regained consciousness after two days, but then deteriorated again and remained comatose for a month. Upon waking up, the individual was amnesiac, nearly blind, had reduced hearing and ataxia. At follow-up (five years later) he had not been able to resume work and evidenced slight motor, memory and visual impairments and clinical indications of slight cerebral atrophy. Electroencephalogram and evoked responses were normal at this time.

Symptoms similar to those described by Tvedt et al. (1991b) were also reported by Kilburn (1993) in an oil well tester rendered semiconscious from exposure to high levels of H2S. Follow-up (39 and 49 months) symptoms suggested damage to the brain stem, basal ganglia, vestibular apparatus, cortex, and other brain structures. There was profound impairment of cognitive function, memory, visual perception and coordination, intelligence and neurophysiologic functions. Malingering was ruled out by the reproducibility of performance during testing sessions. Persistent neurocognitive impairment from H2S poisoning was also reported by Wasch et

al. (1989). Three cases are described with 2/3 individuals experiencing unconsciousness during exposure. All three individuals had moderate to severe attention and concentration deficits, as well as varying degrees and types of memory, learning, and perceptual deficits and prolonged abnormal event-related potentials (P-300 latency). In one individual, symptoms persisted at follow-up three years after exposure.

In another study, a 20-month-old child developed intermittent paroxysmal tonic deviation of the eyes (Gaitonde et al., 1987). After a few months the abnormal eye movements resolved, progressive involuntary movements of the entire body developed, and the child fell frequently. The child was admitted to the hospital with gross truncal ataxia, choreoathetosis, dystonia, and an inability to stand. The child was dysarthic but had normal eye movements. Computer tomography revealed bilateral areas of low attenuation in the basal ganglia and some of the surrounding white matter. Virology was negative and there was no evidence of streptococcal infection. The brain scan suggested toxic encephalopathy. Shortly after admission, the child’s condition improved spontaneously. Ten weeks after admission, the ataxia had resolved and choreathetoid movements were reduced. A repeat brain scan was normal. The child’s illness was attributed to H2S exposure. The family lived next to a coal mine where a burning tip emitted H2S for nearly one year. Three months prior to the child’s admission, H2S emissions were monitored for four months, and the maximum recorded level in family’s housing scheme was 0.6 ppm (0.8 mg/m3). Local authorities admitted that H2S levels may have been many times higher prior to monitoring. The burning tip had been extinguished just prior to the child’s admission to the hospital.

Epidemiological Studies

A mortality study was undertaken among Finnish sulfate mill workers exposed to H2S and organic sulfides (Jappinen and Tola, 1990). Workers had been employed for at least one year between 1945 and 1961 at three pulp and paper mills owned by the same company. No exposure data were presented. Deaths from all causes were not increased. However, workers exposed to H2S and organic sulfides exhibited an increase in cardiovascular-related deaths compared to national death rates (37 observed compared to 24.7 expected). Cardiovascular mortality was higher in workers employed for $ 5 years compared to workers exposed for 1-4 years. The investigators state that increased mortality could not be explained by common risk factors and that differences in smoking habits did not explain the findings. They suggest that increased mortality may have been associated with H2S and organic sulfide exposure.

A cross-sectional study investigated the effects of presumed H2S exposure in sewer workers to determine if chronic exposure to the gas was associated with decreased lung function (Richardson, 1995). Sixty-eight sewer workers performed spirometric tests and results were compared to 60 nonexposed water treatment workers. Job titles were used to categorize sewer workers according to presumed H2S exposure levels. There was a statistically significant decrease in mean FEV1/FVC in sewer workers compared to water treatment workers. The effect was greater in sewer workers presumed to have higher exposures to H2S and longer exposure histories. In nonsmoking subjects, sewer workers were only able to attain 89% of the predicted FEV1/FVC value compared to 98% in water treatment workers. Although the study author states that chronic low level exposure to H2S may be associated with decreased lung function, no measurements of H2S were made so any quantitative relationship between “low levels” of exposure and effect are speculative.

The pulmonary effects resulting from H2S exposure were assessed in 175 workers who extracted and processed oil and gas (Hessel et al., 1997). The workers received a questionnaire concerning sour gas exposures that caused symptoms or loss of consciousness. Thirty-four percent of the respondents indicated that they had exposures serious enough to cause symptoms, and 8% of the workers stated they had experienced a loss of consciousness due to sour gas exposure. In workers that experienced symptoms, no decrease in spirometric values or excess symptoms were noted. While spirometric values were also not affected in workers that lost consciousness, the workers experienced shortness of breath during physical activity, wheezing with tightness in the chest, and wheezing attacks. The investigators stated that these symptoms are consistent with bronchial hyperresponsiveness. The data also demonstrate the refractiveness of the lung to the effects of H2S, with lung function parameters being unchanged despite unconsciousness from H2S exposure.

A cross-sectional pulmonary function study was performed in males that worked in viscose rayon plants (Higashi et al., 1983). A one-workday study was performed in which workers were monitored for H2S exposure for 8 hr and had a forced expiratory flow-volume test performed. Hydrogen sulfide levels were determined using passive diffusion dosimeters worn by the workers (n = 30 in exposed and matched control workers). The workers had been exposed to H2S for an average of 12.3 years. Occupational exposure levels in the exposed workers ranged between 0.3 and 7.8 ppm (2.9 ppm average; 4 mg/m3) compared with < 0.1 ppm (0.1 mg/m3) in matched control workers. No significant differences in the results from the pulmonary function tests were observed in exposed compared to nonexposed workers. Also, no significant differences were noted in the pulmonary function tests taken at the beginning and ending of work shifts.

In addition to the one-workday study, a cross-sectional study was undertaken on 115 exposed workers and 209 nonexposed workers. The subjects underwent forced expiratory flow-volume tests, but were not monitored for exposure. The investigators (Higashi et al., 1983) concluded that chronic exposure to low levels of H2S would not cause adverse pulmonary health effects.

Neurobehavioral effects of H2S were studied in 16 individuals (male and female, ages 21 to 68 years) 2 to 22 years after exposure (Kilburn, 1997). Five individuals had been unconscious after exposure and four had downwind exposures. The exposure level was not reported in any of the cases. Five of the 16 individuals were categorized in the “chronic” exposure group, exposed for 11-22 years. Five were in the “acute” exposure category having been exposed for minutes and the remaining six were in the “hours” category. Twelve of the 16 individuals were plaintiffs in lawsuits. Results from a battery of tests were compared to those from 353 national referents, matched by age, sex, and years of education. Smoking status was not a significant factor. Neurophysiologic and neuropsychologic tests, the latter adjusted for educational level, were employed. Permanent neurobehavioral impairment was apparent in all 16 subjects. Brief high doses were “devastating,” whereas protracted low doses of individuals in the chronic category showed effects (p<0.05) on the more sensitive neurobehavioral tests, such as balance, visual fields, and choice reaction time. There was no indication of bias from the lawsuit and it was considered unlikely that the symptoms could be attributed to other causes. Kilburn (1997) concluded that moderate occupational exposure and downwind environmental exposure can cause permanent impairment. This finding was assessed in a subsequent study.

Neurobehavioral deficits attributed to low-level environmental exposure of 103 individuals to H2S were reported by Kilburn (1999). The neurobehavioral battery included 28 tests. One group of 24 homeowners was exposed to H2S that had collected in crawlspaces and under concrete foundations. Concentrations ranged from 0.1 to 1.0 ppm with several peaks up to 5 ppm during a two-month period in 1996. The homeowners were assessed about one year later. They exhibited abnormal balance, color discrimination, grip strength and delayed verbal recall. It was not stated if the exposure was ongoing at the time of the neurobehavioral assessment. In another group, 48 adults were exposed to measured peak levels ranging from 1 to 20 ppm at street level during a refinery explosion and fire in 1992. They were assessed for neurobehavioral deficits four years later in 1996. The subjects had 11-12 abnormalities compared to 4 in the first group. A third group consisted of 16 patients with a wide range of exposure who were assessed after latent periods of 1.7 to 22 years [these were the same individuals discussed in the preceding paragraph by Kilburn (1997)]. A fourth group consisted of 13 workers and 22 residents living at or near stack gas emissions from a refinery and desulfurization plant. These individuals were also exposed carbon oxide sulfide (2.6 - 52.1 ppm), mercaptans (1.0 - 21.1 ppm), and 24-hr levels of H2S at 0 to 8.8 ppm. Referents (357) were selected from four towns (in four states) free of known chemical contamination. Analysis of variance and covariance adjusted for differences in age, education, gender, and height. Kilburn (1997) concluded that peak dose rather than duration of exposure is predictive of effects, and those exposed to nonlethal levels do not recover completely from H2S. Given the small number (24) of homeowners in the first group exposed to low levels and the type of deficits observed, it is premature to ascribe cause and effect to H2S.