An Exercise in Futility: Speculative Research on Impact of Bhopal Disaster

The study by McCord et al., entitled “Long-term health and human capital effects of in utero exposure to an industrial disaster: a spatial difference-in-differences analysis of the Bhopal gas tragedy” (BMJ, June 13,2023) is reportedly an attempt to provide further insights into the long-term impact of the Bhopal gas leak disaster of December 2-3,1984 on human beings. The study begins by reiterating that: “Evidence from India suggests high numbers of industrial accidents, with the highest associated attributable deaths in the world.”(p.1) It has also correctly pointed out that “…little research is available on…multigenerational impacts, and inadequate recognition of these far-reaching repercussions can diminish the memory of such tragedies for policy consideration.” (pp.1-2) Therefore, the authors claim that: “This research provides insight into these impacts on children born to female survivors of one of the worst industrial disasters in history, the Bhopal gas disaster (BGD).”(p.2)

The main findings of the study were purportedly as follows: (a) “Women who lived within 100 km of Bhopal experienced a relative decrease in the birth of males in the 1985 cohort (64% of children born from 1981 to 1984 were male, a proportion that drops to 60% in 1985)”: and (b) “As in the case of differences in sex ratios, the abnormally high cancer rates appear to extend up to 100 km from Bhopal.” (p.4)

Wrong Assumptions

At the very outset, it may be noted that some of the assumptions with which the study has proceeded are erroneous. For example, the claim that: “The leak spread toxic gas in approximately a 7 km radius around the plant” (p.2) (emphasis added) is an unsubstantiated claim. None of the references provided by the authors in this regard support such an assertion. The toxic plume actually spread to approximately 40 sq km of area outside the pesticide plant due to the breeze that was then blowing in the southeast direction with a velocity of approximately 10-11 km/hr.[1]

The claim that the escaping “MIC toxins affected groundwater” (p.2) is also incorrect. As a matter of fact, groundwater was contaminated due to dumping of toxic waste from the plant – both in pits within the plant as well as in the ‘Solar Evaporation Pond’ (SEP) that was dug adjacent to the plant – while the plant was in operation from 1969 to 1984, i.e., prior to the disaster. Of course, the escaping MIC related toxins did contaminate the Lower Lake and a part of the Upper Lake of Bhopal, which adversely affected aquatic life – at least for a brief period. Unfortunately, the results of numerous studies carried out by State-run institutions like the Indian Institute of Toxicology Research (IITR), Lucknow; Indian Veterinary Research Institute (IVRI), Izatnagar (UP); Veterinary College, Mhow (MP); etc., in this regard have not been made public till date.

Findings Still under Wraps

Findings from several of the Bhopal disaster-related research carried out by institutions like the Defence Research Development Establishment (DRDE), Gwalior; Defence Institute of Physiology & Allied Sciences (DIPAS), Delhi; Institute of Nuclear Medicine & Allied Sciences (INMAS), Delhi; Indian Institute of Chemical Technology (IICT), Hyderabad; National Chemical Laboratory, Pune, etc., are also still under wraps.

Moreover, results of the numerous studies carried out by the Indian Council of Medical Research (ICMR) during 1984-1992 were officially published only 15 to 21 years later, i.e., between 2007 and 2013. From 1994 to 2010, ICMR did not undertake any Bhopal disaster-related medical research. However, due to pressure from victim groups, the ICMR was forced to restart Bhopal disaster-related medical research by establishing the National Institute for Research in Environmental Health (NIREH) in Bhopal in 2010.

NIREH has reportedly conducted a few Bhopal disaster-related studies between 2010 and 2018, the results of which are not yet in the public domain.[2] Since 2018, the only ongoing study related to the Bhopal disaster at NIREH is the ‘Long-Term Epidemiological Survey on Health Effects of Bhopal Toxic Gas Exposure’, which in January 1985 had identified a cohort of about 80,000 in the gas-exposed areas and a cohort of about 16,000 in the purportedly non-exposed areas of Bhopal for carrying out the survey.

Unfortunately, by the time the survey actually began in 1986, the size of the cohorts had come down to about 60,000 in the gas-exposed areas and to about 13,000 in the non-exposed areas. During 1994-2010, when the ICMR had remained practically indifferent to the cause of the Bhopal gas victims, the said survey was kept alive by a team of dedicated doctors under the aegis of the Centre for Rehabilitation Studies (CRS), which was established by the government of Madhya Pradesh with very modest funding from ICMR.

Due to lack of adequate resources and necessary support, the survey team was unable to keep track of the entire cohorts. As a result, by 2010 the size of the cohort drastically came down [see: p.13] to less than 17,000 in the gas-exposed areas and to less than 6,000 in the so-called control areas. By 2023, the size of the cohort has further reduced to just about 10,000 in the gas-exposed areas –COVID-19 providing the additional reason.

No data is available on the fate of the 70,000 gas victims from the original cohort from the gas-exposed areas, who have literally disappeared from the scene. However, even available findings show that morbidity rates in the gas-exposed areas [see pp.87-90] have been consistently high.

Therefore, the said report prepared by CRS & NIREH in 2013 could not but record as follows:

“…it is recommended that newer studies on remaining population of original total gas exposed population of 5,74,000 may be undertaken and extensive follow-up with major focus on clinical disease identification and treatment. The studies should be planned in such a manner so that they can impact guidelines for health service sector to implement treatment measures.” [p.34]

Unfortunately, medical surveillance of the gas-exposed population is still proceeding in a haphazard manner and at an extremely slow pace. It is extremely disheartening to note that during the last 39 years just about 15% of the medical records of individual gas victims have been digitalised and networked so far. The review of the study by McCord et al (2023) is being carried out in this background.

The Problem

It is true that over half a million [official figure of dead and injured is 572,697, see: Para 17, Column G – excluding rows 6 & 7] of the then 900,000+ residents of Bhopal became victims of the disaster and well over 30,000 of them may have perished over the years due to the adverse effects of the toxic fumes. However, according to the figures submitted by the Government of India before the Supreme Court of India, only 5,479 gas victims died and 527,727 others suffered just “minor” injuries (see: Para 17, Column G –  rows 1 & 5).[3]

Nevertheless, the McCord et al. (2023) study under review has noted that:

“There is a broad spectrum of serious long-term and chronic health consequences for hundreds of thousands of survivors, including children, manifesting across multiple systems (eg, respiratory, neurological, musculoskeletal, ophthalmic, endocrine)”. (p.2)

Furthermore, according to the study:

“Potential in utero effects cannot be understated. Studies document a fourfold increase in the rate of miscarriage following the gas leak, as well as increased risk of stillbirth and neonatal mortality. Decades after the disaster, menstrual abnormalities and premature menopause have emerged as common problems among exposed women and their female offspring. MIC has also been shown to damage human chromosomes. Early clinical studies on the gas-affected population showed signs of increased chromosomal aberrations that could manifest as cancer…” (p.2)

Thus, the stated objective of the study was as follows:

“This paper estimates the long-term health (specifically, adult cancer rates and disability) and human capital impacts (educational attainment) on individuals who were exposed to the BGD in utero or as children in 1984.” (p.2)

So far so good!

Questionable Methodology

However, the methodology chosen for conducting the study is highly questionable. While there may be nothing wrong in principle for relying on data generated by the National Family Health Survey (NFHS) or by the National Sample Survey Organization (NSSO), the decision, as stated in the study, to extend the area of coverage of the study for a distance of 250 kms away from Bhopal is totally incomprehensible:

“For the first analysis focusing on differences in the sex ratio, the analytical sample includes children born between 1981 and 1985 to women who were interviewed in Madhya Pradesh and who lived within 250 km of Bhopal (n=1260 births).”(p.2)

What was the rationale for assuming that the impact of the Bhopal disaster may have spread up to a radius of 250 km from Bhopal? It is evident that the said assumption is based on nothing but unwarranted speculation. The inferences from the study are even more startling:

“Given the two spatial patterns in figures 3* and 4**,we use 100 km as the BGD exposure cut-off in our analysis.” (p.5) [Emphasis added]

*In the said study, Figure 3 is interpreted as follows:

“Women who lived within 100 km of Bhopal experienced a relative decrease in the birth of males in the 1985 cohort (64% of children born from 1981 to 1984were male, a proportion that drops to 60% in 1985). This is consistent with male fetuses more affected by external stress.” (p.4)

**Figure 4 is interpreted as follows:

“Similar to patterns seen with changes in the sex ratio, children who were born in 1985 (the year following the BGD) and who lived close to Bhopal during the disaster experience much higher rates of cancer as adults compared with adults who were born before or after the disaster and who lived further away from Bhopal…. As in the case of differences in sex ratios, the abnormally high cancer rates appear to extend up to 100 km from Bhopal.” (p.5)

Erroneous Interpretations

By asserting that “we use 100 km as the BGD exposure cut-off in our analysis”, the authors are trying to establish that the actual impact of the Bhopal disaster was not confined to 40 sq km of area – as assessed by ICMR in 1984, but that its impact had, in fact, spread to no less than 31,416 sq km (an area with 100 km as the radius from the pesticide plant – as claimed in Figure 1 [p.3]), which is nothing but an untenable claim. The authors’ assertion that the area of impact of the Bhopal disaster was 785 times greater than what was assumed previously is based entirely on Figures 3 & 4 – with the causative factor being pinned on the Bhopal disaster, which is highly questionable.

As far as Figure 3 (in the report) is concerned, the reported 4% drop in male births in the 1985 cohort can hardly be considered alarming. Even if the said drop of 4% is assumed to have significance, how does it in anyway prove that the said 4% drop was due to the impact of the Bhopal disaster when there is not an iota of corroborative evidence to substantiate the claim that the toxic fumes or residues had in fact traversed up to 100 km away from the Union Carbide pesticide plant?

The reality is that there was actually an improvement in 1985 in the sex ratio from 64:36 to 60:40 in favour of females, which the authors have tried to interpret negatively. The authors have chosen to rule out the possibility that this development could have been due to reduction in female feticide, which was otherwise rampant. Suppose the sex ratio had further declined to 68:32, would the authors have chosen to interpret the change as a positive development? Moreover, there is no conclusive proof that male foetuses were more adversely affected by external stress. In fact, one of the latest studies on the subject titled “Sexually Dimorphic Accumulation of Persistent Organic Pollutants in Fetuses” (2022) had concluded as follows:

“Prenatal exposure to POPs has been associated with adverse health outcomes, though sex-specific effects are sparse and inconsistent.” [4]