Medical survey of methyl isocyanate gas affected population of Bhopal. Part I. General medical observations 15 weeks following exposure.
The leakage of the deadly gas methyl isocyanate (MIC), from the storage tank of the Union Carbide of India (Limited) factory at Bhopal on the night of December 2-3, 1984 led to the reported affection of over 250,000, people and death of approximately 2000 people The tragedy received wide publicity in the popular press and other media, but no scientific report was published during this period except one on the ophthalmological residual effects. Some of the cases were subsequently referred to the :Department of Respiratory Medicine of the K.E.M. Hospital. As some of the early findings on these patients got reported in the local press and were subsequently pub1ished, we were approached by Nagarik Rahat Aur Punarvas Committee (NRCP), Bhopal, a voluntary health and social welfare group. At their request, we conducted an independent survey on the 16th to 21st March, 1985 i.e. between days 104 to 109 after the tragedy, to determine the extent of the residual damage to different organs in the surviving population, residing in certain areas of Bhopal.
Our team consisted of persons with expertise in internal medicine, respiratory medicine, gastroenterology, nephrology. general surgery, gynaecology, paediatrics. ophthalmology, biochemistry, physiotherapy and laboratory technology in order to carry out haematological, biochemical and pulmonary function tests.
The population covered by the survey was divided into two groups [Fig. I]: Group I had 446 subjects: 114 adult males. 168 adult females and 164 children below 14 years (90 males and 74 females); these subjects resided either (a) within ½ km distance from the Union Carbide factory (Shakti Nagar, Kaichi Chhola, Risaldar Colony, J. P. Nagar), or (b) between ½ to 2 km distance from the factory (Ram Nagar, Rajgarh Colony, Subhash Nagar. Phuta Maqbara, Quazi Camp and Railway Colony). These subgroups (a) and (b) originally studied separately were merged together for data analysis, since no differences were found between them with regard to the degree and the extent of affection. Group II consisted of 123 subjects: 38 adult males, 38 adult females, and 47 children (19 males and 28 females) residing in areas beyond 8 km distance from the factory (Ambedkar Colon). T. T. Nagar and Seva Sadan). Both the groups had similar socio-economic status and resided in similar type of houses. Group II subjects were preselected on these criteria in order to serve as controls for changes likely to be round in Group 1 population. Their participation in the survey was purely voluntary. The survey was conducted in two camps; set up in two different areas-in Kaichi Chhola for the first four days and in Ambedkar colony for the last two days.
The subjects were received and registered by NRPC volunteers, coded and directed to one of the three stations: all males above 14 years (mostly adults) to a station manned by male internists and general surgeons; all females to a second station manned by female internist and general surgeon; and all children below 14 years of age to the third station manned by a paediatrician. A uniform proforma which included the details of history and physical findings was filled for each subject at these stations, and on the basis of the findings they were directed to one or more of the following specialists: ophthalmologist, respiratory specialist, gastroenterologist. gynaecologist and physiotherapist. Each specialist did a detailed clinical survey of the referred case. The physiotherapist did detailed muscle charting in all the cases referred to her and in selected cases performed strength duration curves of both proximal and distal limb muscles on a portable electrodyne. All individuals were requested to undergo urinalysis (macroscopic examination and specific gravity, pH, glucose, ketone bodies, proteins, bilirubin and urobilinogen by multi-dip stick method) and random blood examination for haemoglobin, total and differential WBC counts, blood glucose, total and direct bilirubin, total protein and albumin, SGOT and SGPT, BUN, and serum creatinine. Electrolyte estimation was done using Technicon Autoanalyser on the plasma that was transported to Bombay. Blood samples (Group I-264, Group II-46) were also tested for the levels of thiocyanate, a possible metabolic product of MIC in the body. Blood samples of 50 normal individuals from Bombay, obtained a week later, were also tested for thiocyanate to serve as controls. A few whole blood samples (n = 56) were also flown to Bombay under ice and subjected to carboxyhaemoglobin (COHb) and methemoglobin (MetHb) estimation and blood gases. Forty-one non-smoker healthy subjects were also tested for COHb and MetHb to serve as controls.
In order to obtain an idea of duration: of possible exposure to the leaked gas, specific questions included in the proforma related to the presence or absence of the individual. in Bhopal on the night of December 2/3 and the exact. time of his leaving the area and of return to it. Initial symptoms regarding sensory, respiratory, neuromuscular, ophthalmic and gastro-enterologieal systems were noted. In all adult females, a detailed history of menstrual and obstetric symptoms was obtained. A record was also made of the symptoms present before the gas leakage and of those that were existing at the time of the survey. A few of Group I subjects having persistent gastro-intestinal symptoms were studied in K.E.M. Hospital, Bombay, by upper gastro-intestinal endoscopy and ultrasonography. The data were analysed by Chi Square test and Student's 't' test.
This paper deals with the results obtained in adult subjects. The data on paeditric subjects is dealt elsewhere in this issue. The two groups studied were demographically and socio-economically similar. There were 108 habitual tobacco users among 282 group I adult subjects as compared to 36 out of 76 group II adult subjects.
There were a total of 86 deaths reported in the families of 282 individuals of Group I immediately after the gas leakage as compared to no deaths reported in the families of Group II subjects. This had created a considerable panic in the Group I colonies on the night of gas leakage. Majority of Group I subjects (n = 210, 74.4%) lest their homes within two hours of the leakage, though most of them (n = 134) returned to their homes within 12 hours. whereas the, remaining were either admitted to the hospital for treatment or remained away out of fear of further gas exposure. All of them returned to their localities within 5 days of gas ,leakage. Relevant details of the symptoms experienced by the subjects in both groups immediately after the gas leakage are giver; in [Table - 1]
All except two Group I subjects had symptoms as compared to over 60% of Group II subjects having had no symptom, whatsoever. Similarly, there were. nc critical and severe cases in Group II compared to over 45% of Group I subjects having serious symptoms. Further, far more Group I subjects had significant symptoms referable to eyes and to respiratory, gastro-intestinal and neuromuscular systems as compared to Group II subjects. All these differences were highly significant. Even so, as many as 42% of Group II subjects had eye symptoms and 22.3% had respiratory symptoms; it is therefore clear that even Group II subjects were affected and hence were not. true controls as we had originally expected. based largely on the press reports that residents of Bhopal close to the Union Carbide factor. were the only ones to be affected.
The immediate symptoms experienced by the majority of subjects in both. groups included cough, dyspnoea and chest pain, where as eye symptoms were burning and watering followed by eye swelling. Next in order of appearance were the gastrointestinal symptoms like vomiting, diarrhoea and abdominal pain. However, these occurred mostly in Group I subjects. Neuromuscular symptoms, principally generalised weakness, were the next commonest and were seen only in Group I subjects. Besides these, there were other miscellaneous symptoms like irritation of the skin and dysuria, which were complained of by minority of subjects of Group I (74 of 282) and by only 2 of 76 of Group It subjects.
An assessment. of residual symptoms existing at the time of the survey is shown in [Table - 2]. Group I subjects had significantly higher frequency of persistent respiratory, ophthalmological, neuromuscular and gastro-intestinal symptoms. However, a sizeable number of Group II subjects had similar persistent symptoms. The details of residual gastro-intestinal symptoms seen at the time of survey are shown in [Table - 3]. Of these, the most dominant was upper abdominal pain, coexisting with disturbed appetite and postprandial discomfort.
Neuromuscular symptoms included muscular weakness, which was at times profound and incapacitating. A few subjects complained of a feeling of pins and needles in extremities. Both these symptoms were present in the subjects of both the groups, but Group I subject had much higher frequency. There were no objective neurological signs except in one subject with chronic poliomyelitis. Detailed muscle charting was done in 26 Group I subjects complaining of severe muscular weakness. One of these showed asymmetrical gross weakness, with deformity characteristic of old poliomyelitis. Six others revealed minimal diminution of muscle power, but showed normal strength duration curves in proximal and distal muscles. Inspite of the presence of feeling of pins and needles. none of the subjects showed any objective sensory involvement.
On investigations, no significant abnormalities were observed in the urinary findings of 219 Group I and 47 Group II subjects with regards to the parameters tested, except that three Group I subjects had mild, macroscopic hematuria. The blood counts, liver and renal functions, blood glucose and serum electrolytes were essentially similar in both the groups and were all within normal limits. Blood carboxyhaemoglobin and methaemoglobin levels were considerably higher in Group I Bhopal subjects than in healthy control subjects from Bombay (p< 0.001) [Table - 4]; likewise blood sodium thiocyanate levels were higher in Bhopal subjects (both groups together) than in Bombay healthy controls (p< 0.001), although there were no differences between Group I and Group II subjects.
Information about the toxic effect of methyl isocyanate in humans is scant. Medical personnel treating victims of the Bhopal gas tragedy had no authentic information except the Union Carbide product information booklets,  or the recommendations of the U.S. Department of Health and Human Services and of U.S. Department of Labour. The Union Carbide booklets, , gave no more idea than that "MIC is harmful or fatal if inhaled", "causes burns" and "MIC is harmful if absorbed through skin". The first aid instructions therein refer to (a) removal of the affected to fresh air and providing respiratory support, (b) flushing of eyes and skin with plenty of water or at least 15 minutes, and (c) drinking of plenty of water and induction of vomiting if MIC is swallowed. The U.S. Government information refers "to (a) irritation of the eyes, nose, throat, lungs and skin, and (b) allergy to MIC leading to asthma. Chest skiagrams, FVC and FEV1 may be useful to detect lung damage. Experimental exposure to four human subjects for 1 to 5 minutes caused no effects at 0.4 ppm dose, lacrimation, irritation of the nose and throat at 2 ppm, marked irritation at 4 ppm and unbearable irritations of eyes, nose and throat at 21 ppm. Both these publications did not predict the possibility of tragic consequences as occurred in Bhopal, leading to the death of over 2000, people and according to our estimate disability in about 107,249 people out of about 250,000 exposed.
Adequate toxicity data,, , ,  however, are available in respect of another isocyanate-toluene di-isocyanate (TDI), a polyurethane widely used in the manufacture of urethane. Characteristics of TDI appear to be similar to those of MIC and even the animal toxicity may be regarded to be comparable. Human volunteers can recognise the characteristic odour of TDI at concentration of 0.05 PPM, but not at concentrations of 0.01-0.02 ppm. At higher concentrations of 0.05-01 ppm eye, nose and throat irritation occurred in decreasing frequency. Still higher concentrations lead to universal irritation. One study refers to more than 100 severe illnesses, quoting 4 deaths caused by TDI. In an excellent summary of all authentic available world literature until then, 234 mild cases, 78 severe cases and 2 deaths to be due to TDI toxicity were mentioned. However, no deaths due to MIC were reported before Bhopal disaster, but a warning about fatal hazard from isocyanates as a group was indeed available from TDI data. The concentration of MIC on the night of the Bhopal tragedy must have exceeded several times 21 ppm, which was recorded to cause only unbearable irritation of eyes, nose and throat.
Immediate effects of MIC gas leakage in our subjects were discernible in both groups. However, there were some important differences in their manifestations. Group II subjects had only mild to moderate symptoms, no neuromuscular symptoms and no deaths in their families. Only one subject had gastro-intestinal symptoms; respiratory symptoms were present in 17% and the eye symptoms were the most common but present in only 40% of subjects. In comparison, Group I subjects had severe and critical disease in 45%, mostly unconsciousness and had 86 deaths among their family members. The respiratory and eye involvements were nearly always present and about half of them had respiratory and neuromuscular symptoms. The differences point to the possibility that an extremely high concentration of MIC in Group I areas caused lethal consequences in many and disastrous effects in those who survived. Group II areas, originally thought to be control areas, being far away from leakage, had unmistakable evidence of immediate MIC exposure. The concentrations of MIC in air in Group II areas soon after the leakage must have been somewhere between 4 and 21 ppm. On enquiry, it was also revealed that residents of intervening areas of Bhopal too had variable effects of gas exposure. We based our sampling on the newspaper reports indicating that residents of nearby areas alone were involved. However, to date no epidemiological study of Bhopal gas leakage is available.
A survey of the delayed effects of MIC exposure revealed that in both the groups considerable disability existed, comprising respiratory, ophthalmic, gastro-intestinal and neuromuscular symptoms. Group I subjects had much greater involvement than Group II. Group II subjects had mild to moderate symptoms which were not disabling or crippling, but were persistent in about 75% of subjects, indicating chronic damage in this group, which cannot therefore be accepted as a background population system. Further studies should take into consideration a control population which is far away to have any effects of MIC exposure.
The respiratory, ophthalmic, gynaecological and obstetric changes are dealt with elsewhere in this issue. These were the most prominent symptoms. However. gastro-intestinal and neuromuscular symptoms also contributed significantly to the overall morbidity. Analysis of gastro-intestinal symptoms strongly indicates the involvement of the upper gastro-intestinal tract, gall bladder or the pancreas. No physical signs were present. The non-availability of further investigative facilities prevented us from drawing any precise conclusions about the source of these complaints, but gastroduodenal involvement seems most likely. Upper gastro-intestinal endoscopy done on a small number of subjects from Group I admitted to our hospital 7 to 9 months after exposure revealed no gastroduodenal ulceration, but only evidence of non-specific antral gastritis or duodenitis; and ultrasonography showed normal structure of the gall bladder and pancrease. No studies of pancreatic functions were possible, but pancreatic damage may take several years to be established before it is detected by pancreatic function tests. The most likely source of these complaints. therefore, seems to be non-ulcer dyspepsia. an ill-understood term, often shown to be associated with histological evidence of gastroduodenitis. No studies on the effects of isocyanates on gastric secretion are available, but thiocyanates are known to be inhibitors of gastric acid secretion, by forming permanent hydrothiocyanic acid within the acid secreting cells and increasing the rate of protein loss. A few subjects showed evidence of functional disturbances of large bowel in the nature of frequent small volume stools. The chronicity of these symptoms precluded the possibility of any parasitic or bacterial infection and these cases could be conveniently labelled as irritable bowel syndrome on clinical grounds. No subject had any evidence of small bowel involvement. which if present, would have indicated organic damage. The overall gastro-intestinal symptoms in these cases point to functional involvement which may either be related to non-specific inflammation of the gastro-intestinal tract of uncertain significance or may be due to psychosomatic causes, which were possibly operating in our subjects.
Neuromuscular symptoms present subjects of both groups also need an explanation as regards their causation. The weakness was often profound and incapacitating in Group I subjects, but objectively there was no more than minimal loss of muscle power in the six subjects tested in detail. The absence of objective sensory signs was also observed in subjects who complained of feeling of pins and needle;. The overall picture is difficult to explain, but we were impressed by the persistence of symptoms. Some patients with muscular weakness but without objective neurological signs have been described to have electromyographic changes only on supramaximal exertion. Mitochondrial damage has been shown to occur due to deficiency of cytoehrome oxidase and other respiratory chain enzymes. Since cyanides and thiocyanates, the metabolic products of isocyanates in the body, are known to be inhibitors of these enzyme systems, it can be postulated that our patients with severe subjective weakness but without objective proof might be having such damage.
Investigations testified to the lack of persistent affection of organs like the kidney, the liver or the hemopoietic system. However, blood carboxyhaemoglobin and methaemoglobin levels as well as blood thiocyanate levels in Group I subjects were significantly higher than in the controls. Blood thiocyanate levels among a few of our team members were also found higher than in normals and returned to normal after 7 weeks' period away from Bhopal. We have observed that thiocyanate levels of water from Bhopal lakes continued to be high at the time of our survey and when tested after a further period of 8 weeks showed a considerable decline. The thiocyanate levels of tap water in Bhopal also showed similar levels as lake water at this stage. Even at this point, however, the levels were twice as high as in water supplied for domestic consumption in Bombay. All this evidence suggests that thiocyanate existed in the Bhopal environment even 160 days after exposure although gradually the levels are falling. At the end of 9 months, thiocyanate levels in tap water and lake water had returned to normal. These thiocyanates might have contributed to the persistent morbidity, although this contention needs more substantiation. Moreover, it is unlikely that the persistence of increased levels of altered blood haemoglobin and thiocyanate levels is a result of exposure at the time of the gas leakage in December 1984. We believe that this could be attributed to continued consumption of water and food contaminated by breakdown products of MIC viz. cyanides and thiocyanates.