Hindsight check: LL2-05 (Ch 5 Minamata disease: a challenge for democracy and justice)#

Source section: EEA, Late lessons from early warnings: science, precaution, innovation (EEA Report No 1/2013), Ch 5, report pp. 92–130. Chapter by Yorifuji, Tsuda and Harada (pp. 92–120); panels by Grandjean (pp. 121–124), Selin (pp. 125–129) and Castaño (p. 130). Check window: publication (early 2013) to late September 2026. Checked: 25 September 2026.

Method note. - Search was unavailable. The session’s general web-search budget was already used up, so I could not run open searches. Instead I fetched known primary repositories directly and used their own indexes: - Japanese government sources: - the Ministry of the Environment (MOE): its Minamata disease policy page, press releases, and verbatim transcripts of 160+ ministerial press conferences from October 2024 to 15 September 2026; - Kumamoto, Kagoshima and Niigata prefectures; - the Minamata City Municipal Museum; - the Supreme Court judgment database. - International bodies: the UN Treaty Collection, the AMAP/UNEP Global Mercury Assessment 2018 technical report, and the US EPA’s IRIS database (Integrated Risk Information System). - Journals: the Europe PMC and Crossref bibliographic databases, for peer-reviewed abstracts. - Pages that refused automated retrieval (HTTP 403): the Minamata Convention Secretariat site (minamataconvention.org), the UNEP website and document repository, IISD’s Earth Negotiations Bulletin (ENB), and Mainichi. Several Japan Times wire stories from 2023–2025 have expired. - News sources. Where I rely on news, it is the Japan Times (English) or the Asahi Shimbun (Japanese; paywalled, so lead paragraphs only), and I say so. I had no general news search, so a court ruling or announcement reported only in the press could have been missed. I flag this where it matters, mainly for the appeals pending in September 2026. - Annex 3 does not apply. Minamata was not one of the 2001 volume’s cases, so Annex 3 of the 2013 report has no update to note. - Strand boundary. I read only the digest and the source extract for this section, plus web sources.

Overview#

Thirteen years on, the chapter’s core diagnosis of the Japanese case has held up remarkably well. In several places events since 2013 have strengthened it. Its framing needs one legal correction, and the panels’ global forecasts have fared unevenly.

What held or strengthened (Japan). - Recognition is frozen. Official recognition has barely moved. - Kumamoto and Kagoshima had 2,273 certified patients in September 2011 and 2,284 in 2022. Kumamoto’s count of 1,791 has not changed between April 2022 and August 2026. - Nationally, about 3,000 people are certified, including 717 in Niigata. - Meanwhile the 2009 Special Measures Act brought 55,000 people into relief without recognising them as patients. That is itself an official measure of the “several tens of thousands” the chapter described. - The 1977 criteria are unchanged. - The Supreme Court (16 April 2013) held that they are not the only route to recognition. - MOE responded with a 2014 notice on “comprehensive assessment”. Critics said it added new evidential burdens, and certifications stayed in single digits. - In March 2026 the Niigata District Court ordered certification of all eight plaintiffs whose applications had been rejected, as reported at the minister’s press conference. The prefecture appealed after MOE warned the ruling could disrupt “uniform operation” of the certification system. - In April 2026 the Fukuoka High Court went the other way and upheld Kumamoto’s and Kagoshima’s rejections. - Litigation continues, as the chapter predicted. - After 2013 roughly 1,800 people who had not received relief under the 2009 Act sued, in four courts. - The first-instance results are split: - Osaka (2023) recognised all 128 plaintiffs as having Minamata disease and awarded damages. - Kumamoto (2024) found 25 of 144 had the disease, but ruled their claims time-barred. - A Niigata ruling in April 2024 awarded damages against the polluting firm. I saw only a headline for this ruling. - The appeals had not been decided by mid-September 2026, as far as I could find. - No population survey yet. The residents’ health survey that Article 37 of the 2009 Act told the government to carry out “positively and promptly” has still not been done. - The only fieldwork so far is a feasibility study: 32 participants, November 2025 to January 2026. - The main survey is due to start in fiscal 2026. After a July 2026 earthquake, its timing and methods were under review. - The ministry frames its purpose as “dispelling health anxiety”, not finding unrecognised victims. - Episodes of poor process. Several recent episodes fit the chapter’s “democratic deficit” and secrecy diagnosis: - officials cut victims’ microphones at a May 2024 meeting with the minister, following a pre-written script; - the government’s environmental White Paper counted only living certified patients for about 30 years until 2026; - the ministry says a third political settlement lacks “material” to justify it.

Balancing evidence: the ministry apologised, lengthened its dialogues with victims’ groups, and in April 2026 linked relief allowances to inflation.

What needs correcting or qualifying. - The Supreme Court did not “declare invalid” the 1977 criteria (pp. 92, 94). - The chapter’s own chronology says the 2004 court “ruled that the 1977 criteria should be relaxed” (p. 115). - The 2013 judgment, which I read, calls the criteria rational “to that extent”, as a device for rapid inference. It says they do not exclude individual recognition by other evidence. - The government’s line that no Supreme Court ruling has rejected the criteria is therefore legally accurate, even if it misses the point. - The claim that in-utero exposure falls outside the 2009 Act is partly inaccurate (p. 110). The 2010 relief policy expressly considered fetal exposure for people born up to November 1969. The stronger version of the chapter’s point survives: the Act’s symptom tests are built around sensory disturbance and may miss cognitive effects of in-utero exposure. The authors have since published data on exactly those cognitive effects. - The 488-tonne discharge figure (p. 105) is at the high end. Official Japanese sources cite 70–150 t “or more” deposited in Minamata Bay, which is a different quantity. I could not check the original source. - The diagnostic dispute is still live in the scientific literature. - Government-aligned experts accused victims’ doctors of “mass biases” in their surveys in a 2024–25 exchange. - Two courts (Kumamoto 2024, Fukuoka High Court 2026) accepted the state’s arguments about distance from the source and latency.

The chapter’s framing is well supported by the Japanese Society of Psychiatry and Neurology (JSPN) and several courts. It is still one side of a contested medical question, argued by protagonists.

The panels’ global claims. - The Minamata Convention (Claim 5): held up as a plan; unclear on effectiveness. - It was adopted in Kumamoto on 10 October 2013 and entered into force on 16 August 2017. It had 154 Parties at 26 September 2026. - In November 2025 the Parties agreed to end dental amalgam by 2034. - The first formal effectiveness evaluation is still under way. - Global emissions have not yet fallen. One 2025 reconstruction finds they “continued to rise slightly” after 2013 because growth in the Global South offset declines elsewhere. - Tropical tuna mercury has been flat since 1971. - Selin on Asian emissions (Claim 7): partly held up. - China’s emissions rose to about 530 t/yr in 2014 and have declined since the 2010s, so the “dramatic increase” from China did not happen. - Growth shifted to other Global South countries. - Artisanal and small-scale gold mining (ASGM) remains the largest source. - A 2016 study also challenged Selin’s premise that global emissions were flat after 1990. - Selin on long timescales and fast local responses (Claim 8): held up and reinforced. Evidence includes ocean inertia in tuna, continued declines in Massachusetts fish, and modelling of the costs of delay. - Grandjean’s call to cut the exposure limit by about 50% (Claim 6): contested and unadopted. - The US EPA reference dose for methylmercury, 0.1 µg/kg/day, dates from 2001. A reassessment has sat at “step 1” since 2019. - The European Food Safety Authority (EFSA) reaffirmed its 2012 tolerable weekly intake of 1.3 µg/kg bw in June 2026, while noting an update “might merit”. - The low-dose evidence is mixed: - an EPA-authored 2026 meta-analysis finds a steeper IQ slope at lower exposures, but the confidence interval includes zero; - a 2022 systematic review found only weak evidence of harm at dietary levels. - Castaño’s DEMOCOPHES pilot (Claim 9): held up. It reported in 2013–2015. Harmonised human biomonitoring was then institutionalised in the European Human Biomonitoring Initiative (HBM4EU) and is now used for Convention effectiveness evaluation.

Implication for weight. - Mechanisms that carry strong weight. On the Japanese case, the chapter’s mechanism-level lessons now rest on a 13-year out-of-sample record, not just on the authors’ history: - hardened case definitions; - settlement without recognition; - passive, claim-based counting; - the state as co-financier of the polluter’s liabilities; - procedural disrespect toward claimants. - Claims that carry less weight: - its legal characterisations; - its single-sourced numbers; - Grandjean’s claim about exposure limits, which remains a respected but minority reading.

The global-emissions forecasts were directionally right about where the problem would move (the Global South, gold mining). They were wrong about China.

Claim-by-claim#

Claim 1: 2,273 recognised patients (September 2011) versus “several tens of thousands” with characteristic symptoms unrecognised and not properly compensated (pp. 94, 110)#

Original claim. - “In September 2011 although 2 273 individuals were officially recognised as Minamata Disease patients… several tens of thousands have neurological symptoms characteristic of methylmercury poisoning but remain formally unrecognised” (p. 94). - The same point is repeated at p. 110. The count covers Kumamoto and Kagoshima, the Shiranui (Yatsushiro) Sea area.

Subsequent developments. - Certified patients have barely changed. - Minamata City Municipal Museum, April 2022: 2,284 certified, of whom 1,791 are in Kumamoto and 493 in Kagoshima. That is 11 more than in 2011. The museum’s year-by-year table shows 3 certifications in 2013, 1 in 2014, 3 in 2015, 2 in 2016, 1 in 2019 and 1 in 2022. Of the 2,284, 2,024 had died by August 2022; 268 were alive in April 2022. (Minamata disease: its history and lessons 2022, Jan 2023) - Kumamoto Prefecture, end of August 2026: still 1,791 certified, against 13,569 rejections and 217 applications pending. (Kumamoto Prefecture, certification status as of end Aug 2026, page updated 10 Sep 2026) - Kyodo, via the Japan Times, 1 May 2025: still 2,284 for the two prefectures. (Japan Times tag listing, 1 May 2025; article expired, lead visible) - Niigata, end of July 2026: 717 certified (377 by the prefecture, 340 by Niigata city) and 1,679 rejected. (Niigata Prefecture press release, 10 Aug 2026) - The Environment Minister gave the national total as “about 3,000” on 10 July 2026. (MOE press conference, 10 Jul 2026) - Relief without recognition dwarfs certification. Applications under the 2009 Special Measures Act closed on 31 July 2012, and determinations were completed in January 2018. (MOE, results of relief determinations) Results across the three prefectures:

• The minister summarised the position as about 3,000 compensated under the Pollution-related Health Damage Compensation Act, plus “more than about 50,000” covered by the 1995 and 2009 political settlements. (MOE press conference, 10 Jul 2026)
• Official presentation of the numbers changed only in 2026. For about 30 years the Environment White Paper listed only living certified patients. The 2026 White Paper (Cabinet decision 5 June 2026) added a table including the deceased, after requests from victims’ groups. The minister said the earlier presentation was “not inappropriate”. (MOE press conference, 5 Jun 2026)
• Estimates above “tens of thousands” are advocacy figures. A 2017 Japan Times editorial reported doctors’ estimates of 100,000–200,000 victims (21 Jan 2017). These come from victims’ physicians and are not an official figure.

Verdict: held up. The recognised count is essentially frozen, and official relief figures (about 55,000 under the 2009 Act alone) confirm an unrecognised symptomatic population of tens of thousands. - Caveat 1: relief is not diagnosis. Relief eligibility rests on exposure history and sensory disturbance, not a clinical diagnosis. The government treats recipients as “victims” warranting relief, not as patients. - Caveat 2: the courts are split. Courts disagree on how many such claimants have Minamata disease (see Claim 3).

Implication for weight. Lesson 13 in the digest carries strong weight: “settlement without recognition prolongs disputes; passive, claim-based surveillance hides scale”. A case count that only grows when individuals apply, and that the payer adjudicates, stayed flat for 13 years. Meanwhile an administratively defined relief population 20 times larger was paid off-register.

Claim 2: The 1977 criteria, judged medically invalid by JSPN (1998) and ruled by the Supreme Court (2004) to need relaxing, “remain in force today”; the government has not changed them (pp. 92, 94, 109–110)#

Original claim. - The criteria “were judged medically invalid by the Japanese Society of Psychiatry and Neurology (JSPN) in 1998 and declared invalid by the Supreme Court in 2004… Nevertheless, the government has not changed the criteria” (p. 94). - The summary on p. 92 says the Court “declared the definition invalid”. The chronology on p. 115 says the Court “ruled that the 1977 criteria should be relaxed”.

Subsequent developments. - Supreme Court, 16 April 2013. - Kumamoto case: the Court upheld the Fukuoka High Court’s order to certify the late Chie Mizoguchi, who had applied in 1974. - Osaka case: it reversed the Osaka High Court. - The leading judgment (Heisei 24 (Gyo-Hi) No. 245) holds that courts must judge each applicant individually, “comprehensively examining the circumstances and relevant evidence in light of rules of experience”. Courts should not merely ask whether the administrative criteria are unreasonable. - On the criteria themselves, the Court said they have rationality “to that extent”, as a standard allowing rapid decisions by inference from general knowledge. But they “cannot be said to exclude” certification where the listed combination of symptoms is absent. - Kumamoto then withdrew its remanded appeal (7 May 2013). (Supreme Court judgment summary and full text; companion case; Kumamoto Prefecture chronology 1956–FY2025) - MOE notice, 7 March 2014. MOE issued a notice on how to conduct the “comprehensive assessment”. - The Japan Times editorial board judged that it “appears to relax conditions… but this is not the case”, because it demands evidence of exposure. (17 Mar 2014) - Japanese Wikipedia (secondary) summarises the notice’s exposure indicators. These include certified patients in the same household or area, a fishing history, and hair-mercury data from the period of contaminated fish consumption. - Certification figures since 2014 (Claim 1) are consistent with the critics’ reading. - The government’s position in 2026. On 24 March 2026 the minister said “previous Supreme Court rulings have not rejected the current certification standard, the 1977 criteria”. He said the 2014 notice gave concrete form to the 2013 ruling’s call for comprehensive assessment, and that the ministry maintains the 1977 criteria. On 13 March 2026 he said he had no plan to review the certification system. (MOE press conference, 24 Mar 2026; 13 Mar 2026) - Courts diverge on applying the criteria. - Niigata District Court, 12 March 2026: it overturned rejections and, per the questioning reporter, ordered certification of all eight plaintiffs. - Niigata Prefecture appealed on 26 March 2026. It explained that MOE had expressed “strong concern” that the ruling adopted “a judgement framework different from past rulings of the same kind, including in higher courts”. MOE also warned it would greatly affect “uniform operation” of certification if it became final. (Niigata governor’s comment on the ruling, 12 Mar 2026; appeal statement, 26 Mar 2026; MOE press conference, 13 Mar 2026) - Fukuoka High Court, 23 April 2026: it dismissed an appeal in an administrative suit, finding for Kumamoto and Kagoshima. The plaintiffs announced an appeal to the Supreme Court. This appears to be (my inference from the parties named) the 2015 suit by members of a mutual-aid society of Minamata disease victims seeking mandatory certification, which the Kumamoto District Court rejected on 30 March 2022. (MOE press conference, 28 Apr 2026; Kumamoto chronology) - A 2014 suit seeking to block the 2014 notice was dismissed as inadmissible; the Supreme Court declined review on 1 December 2015 (Kumamoto chronology). - The administering authority asks for change. - Niigata carries out certification as a function delegated by the state. Its governor again asked on 9 June 2026 for the standards to be reviewed. He also said changing standards would take time and a political solution was needed. (MOE press conference, 12 Jun 2026) - The Niigata city mayor called the ministry’s response “as if it were someone else’s problem” (MOE press conference, 24 Mar 2026). - The medical dispute continues in print. In 2024–2025 three senior figures (Futatsuka, Eto and Uchino) reviewed the victims’ physician Takaoka’s book in the Japanese Journal of Hygiene. They accused his side of “mass biases in epidemiological surveys” and toxicological errors. Takaoka replied that the reviewers’ logic confirmed the “inaction of experts”. (Takaoka 2024; Futatsuka et al. 2025; Takaoka 2025)

Verdict: held up, with a correction to the chapter’s legal wording. - What held up: - The criteria remain formally in force, unchanged since 1977. - The government has repeatedly declined to revise them. - Its main response to adverse rulings has been to defend uniform application. - The correction: the Supreme Court has never “declared invalid” the 1977 criteria. The 2013 ruling was the strongest move against them, and it made the criteria non-exclusive rather than invalid.

Implication for weight. The “hardened case definition” mechanism carries strong weight (digest insight 11). The post-2013 record shows the mechanism at work in real time: - a definition set under cost pressure survives professional-society censure; - it survives a top-court ruling; - it survives a delegated administrator’s requests for review; - the central ministry’s appeal rationale is administrative consistency, not medical validity.

Users should cite the legal history precisely. Overstating it (“declared invalid”) hands critics an easy rebuttal.

Claim 3: The 2009 Act (JPY 2.1 million relief without liability or patient status) was expected to end litigation, but lawsuits continue because areas and periods were defined without investigation, and in-womb-exposed people who are not congenital patients are excluded (pp. 110, 115)#

Original claim. - “It is expected that most of the applicants involved in litigation will withdraw their actions… several lawsuits are still under way because the Act defined the affected area and time period without investigating or defining evidential criteria” (p. 110). - Individuals “exposed in the uterus but… not severely enough affected to be recognised as ‘congenital’ Minamata disease patients are not covered by the Act” (p. 110).

Subsequent developments. - A new wave of mass litigation followed the Act. - People with the Shiranui Patients’ Association (Minamata disease victims not relieved under the Act) filed suits under the “No More Minamata” banner, the second such round: - Kumamoto District Court, June 2013; - Tokyo, August 2014 and April 2017; - Osaka, September 2014. - A separate Niigata series began in December 2013. (Kumamoto chronology; Soshisha litigation list, updated June 2022 with entries to 2024) - By 2024 nearly 1,800 people had filed similar suits, 1,400 of them in Kumamoto (Japan Times tag listing, 22 Mar 2024; Asahi, 22 Mar 2024). - First-instance results are split: - Osaka District Court, 27 September 2023. It recognised all 128 plaintiffs as having Minamata disease and ordered the state, Kumamoto Prefecture and Chisso to pay ¥2.75 million each. The state and prefecture appealed to the Osaka High Court on 10 October 2023. (Japan Times tag listing, 27 Sep 2023; Kumamoto chronology) - Kumamoto District Court, 22 March 2024. It dismissed all 144 plaintiffs. - It found that 25 had Minamata disease but applied the 20-year limitation period, which Asahi called unusual in Minamata cases. - It adopted the state’s positions: - sensory findings depend on the patient’s “subjective” report; - exposure falls off with distance from the source; - onset occurs within “roughly 10 years” of the end of pollution. - The plaintiffs’ lead lawyer argued that recognising 25 as having the disease showed the Act’s relief had “errors”. The plaintiffs appealed to the Fukuoka High Court on 4 April 2024. (Asahi, 22 Mar 2024; Kumamoto chronology) - Niigata District Court, April 2024. It ordered damages over Niigata Minamata disease. I saw only the headline; the Kyodo text has expired, so I cannot give figures. (Japan Times tag listing, 19 Apr 2024) - Appeals remain undecided, as far as I could find. The official Kumamoto chronology (to March 2026) and the minister’s press conferences (to 15 September 2026) record no appellate judgment in the Osaka, Fukuoka or Tokyo No More Minamata appeals. New suits are still being filed, for example in Tokyo on 16 March 2026 (Kumamoto chronology). - Political response. - Five opposition groups re-submitted a relief-expansion bill on 9 July 2026. A governing-party committee instead asked for more medical and welfare funding. - After meeting victims on 1 May 2026, the minister said he could not find “material” to overcome the history and justify a third political settlement. He did not see public understanding for one. (MOE press conference, 10 Jul 2026) - From April 2026, monthly medical allowances rose by ¥1,500 for people relieved under the 2009 settlement and by ¥1,400 for people relieved under the 1995 settlement. The 2009 allowances will also be adjusted for inflation annually from fiscal 2027. (MOE press release, 21 Apr 2026; minister’s message, 26 Dec 2025) - On in-utero exposure, the chapter is partly inaccurate. The Cabinet’s April 2010 relief policy under the Act says: - For Kumamoto and Kagoshima, people born by the end of November 1969 are to be assessed “taking into account the possibility of fetal exposure”, together with the area and symptom requirements. - People born later can qualify with data such as umbilical-cord or maternal-hair mercury.

(MOE, relief policy, Cabinet decision 16 Apr 2010) So in-utero exposure was not categorically excluded. The defensible version of the chapter’s point is that the symptom requirements centre on sensory disturbance of the extremities. Moderate prenatal exposure mainly produces cognitive deficits (see Claim 4), so many people exposed in the womb may fall outside the Act in practice.

Verdict: held up on the central prediction: the Act did not end litigation, and the area and period boundaries are the crux of the dispute. The subsidiary claim about in-utero exclusion is partly inaccurate as a statement of the relief policy.

Implication for weight. There have been two “final and complete solutions” (1995 and 2009), each paying relief without admitting liability and each followed by a new litigation wave. That is strong evidence for the digest’s “false closure” mechanism (insights 8 and 13). The split court record means users should not claim the courts have vindicated the victims across the board. They have not, and the state’s scientific arguments about distance from the source and latency have won in two courts.

Claim 4: Affected areas and communities have never been properly investigated; follow-up epidemiology of congenital and moderate in-womb exposure is needed (pp. 92, 106–107, 110)#

Original claim. - “The continuing failure to investigate which areas and communities were affected means that the financial settlement’s geographic and temporal scope is still not properly determined” (p. 92). - Further investigation of moderate in-utero effects is needed (pp. 106–107).

Subsequent developments. - The legal duty has existed since 2009. Article 37 of the 2009 Act requires the government to “positively and promptly” study the health of residents of the designated and surrounding areas, “including those who were fetuses when Minamata disease was frequent”, and to publish the results (Act text). - Timeline of the government survey: - September 2020: MOE said it would organise results on methods for objectively measuring methylmercury effects within one to two years (Kumamoto chronology). - July 2024: after the microphone incident, Minister Ito pledged to be ready to start a health survey within about two years (MOE press conference, 4 Oct 2024). - December 2024 – March 2025: an expert panel met in closed session and endorsed interviews, neurological examination, magnetoencephalography (MEG, a brain-scanning technique) and MRI. - Victims’ groups objected to the process and the method. - A physician they nominated was heard, and raised concerns about the burden on participants. - (MOE press conference, 28 Jan 2025; 3 Oct 2025) - MOE research group, fiscal 2024: eligibility excludes people already recognised or relieved, and those with diabetes, heavy drinking, metal implants or other neurological disease (MEG/MRI study group report summary). - Feasibility study: - Invitations went in October 2025 to 800 residents of Amakusa and Kami-Amakusa born before 1975, drawn at random from the resident register. - 32 took part (November 2025 – January 2026). - Of 44 non-participants who gave a reason, 29 met exclusion criteria such as diabetes or metal implants. - (MOE announcement, 1 Oct 2025; preliminary results, Feb 2026) - April 2026: the main survey of Shiranui Sea coastal residents was to start “as early as this year” (Japan Times, 30 Apr 2026). - 15 September 2026: MOE said timing and methods would take account of the 28 July 2026 Kumamoto earthquake, but it still aims to start within fiscal 2026 (Asahi, 15 Sep 2026). - The stated purpose excludes case-finding. Asked whether the survey aims to uncover unrecognised victims, the minister said its purpose is “basically dispelling health anxiety in the region”. He said “digging up” victims may not be the right term, and the aim is understanding “the local situation” (MOE press conference, 3 Oct 2025). A reporter noted that victims’ groups say no comprehensive regional damage survey has been done in 70 years. - Independent studies have filled some of the gap. Most are by the chapter’s lead author or allied clinicians: - Yorifuji et al. (2025) reanalysed 1970 data on 416 junior-high students. Neurocognitive impairment was 2.08 times more prevalent in the highly exposed Minamata area than in Amakusa (95% CI 1.26–3.44). (Sci Total Environ 2025) - Yorifuji et al. (2026) examined people now about 70 who were prenatally exposed but are not congenital patients. They scored 5.91 points lower on the Japanese Montreal Cognitive Assessment (MoCA-J) than less-exposed controls (n = 51). (Neurotoxicol Teratol 2026) - Yorifuji, Takaoka and Grandjean (2018) found that congenital patients’ daily functioning declined faster than expected with age over 10 years (n = 11) (Neurotoxicol Teratol 2018). - Takaoka et al. (2024) found limb-predominant sensory disturbance in 38–59% of residents of three distant Yatsushiro Sea districts, against 1.4% in an unexposed control area (Toxics 2024). - Their methods are contested. See the Futatsuka–Takaoka exchange under Claim 2. - Legacy containment was tested. The 28 July 2026 earthquake liquefied parts of the reclaimed land in Minamata Bay where mercury sludge was contained in 1990. Kumamoto reports no mercury detected in the ejected soil; victims’ groups want stronger measures. (Asahi, 16 Sep 2026)

Verdict: held up, and partly strengthened. - Held up: as of September 2026 no population-based health survey of the affected region has been carried out, 17 years after a statute required one “promptly”. The planned survey is framed as reassurance, not case-finding. - Partly strengthened: the chapter authors’ own follow-up studies report the moderate in-utero effects they predicted. These are small studies by interested researchers, and not independent confirmation.

Implication for weight. “Passive surveillance hides scale” (digest insight 13) and “the uterus is part of the environment” (p. 113) carry strong weight. The post-2013 record also adds a technology-neutral pattern: a statutory duty to investigate can be discharged slowly, through method development, closed expert panels and feasibility pilots. Meanwhile the authority conducting the survey defines its purpose narrowly enough that it cannot reopen settled boundaries.

Claim 5: Japan proposed the name “Minamata Convention” and to host adoption; signature was planned in Japan in 2013 (pp. 93, 115)#

Original claim. - Japan’s delegate proposed the name to signal that the Minamata disaster “would never be repeated”, and offered to host the conference (p. 93). - Signature was “planned for signature in Japan in 2013” (p. 115). - Harada was sceptical of Japan’s claim to be a model (pp. 93–94).

Subsequent developments. - Adoption. The Convention was adopted unanimously at a diplomatic conference in Kumamoto and Minamata on 10 October 2013. Ninety-two signatories, including the EU, signed there. (MOE, Convention overview) - Reception in Japan was contested. - Victims objected to the treaty’s name, and critics said it had “too many loopholes” (Japan Times, 7 Oct 2013). - Prime Minister Abe “angered Minamata disease victims” by saying that Japan had “recovered” from Minamata disease (Japan Times, 21 Oct 2013). This vindicates Harada’s scepticism on p. 93. - Entry into force and ratification. Japan ratified on 2 February 2016. The Convention entered into force on 16 August 2017. At 26 September 2026 it had 128 signatories and 154 Parties. The depositary gave notice that amendments had entered into force on 5 October 2023 (Annex A) and 5 May 2025 (Annexes A and B). A further Annex A amendment was notified on 23 March 2026. (UN Treaty Collection status page) - COP-6 (Geneva, 3–7 November 2025). The Parties: - added dental amalgam to Annex A with a 2034 phase-out date; - launched work on regulating trade in mercury compounds; - agreed international cooperation on gold supply-chain traceability for ASGM; - deferred a decision on mercury-free vinyl chloride monomer production to COP-7 (June 2027).

(MOE press release on COP-6 results, 13 Nov 2025) - Effectiveness evaluation (Article 22) is still in progress. - A scientific group (the Open-Ended Scientific Group, OESG) and a governmental evaluation group are compiling it. The OESG’s second meeting was held in Minamata on 17–21 March 2025 (MOE press release, 11 Mar 2025). - The COP-6 summary reports no effectiveness conclusion. - Early indicators of effect are weak. - Qiu et al. (2025) find that global emissions “have continued to rise slightly since the 2013 Minamata Convention”. Declines in the Global North and China have been more than offset by the rest of the Global South. (Nat Commun 2025) - Tropical tuna mercury shows “stable long-term trends” from 1971 to 2022, attributed to ocean inertia (Médieu et al., ES&T Letters 2024). - One national case shows a quick exposure response to product phase-outs: Slovenian children’s urinary mercury fell from 0.76 to 0.22 ng/mL (2007 to 2018–24) as dental amalgam use fell (Usenik et al. 2026).

Verdict: held up on the factual forecast: named, hosted and signed in Japan in 2013. The Convention’s effectiveness is unclear: it is ratified, in force and tightening, but global emissions have not yet turned down, and the formal evaluation is incomplete.

Implication for weight. The chapter’s framing contrasts international leadership with unresolved harm at home. The events of October 2013 bore it out. More broadly, the story shows that a successful international regime can coexist with, and be used rhetorically to paper over, unresolved domestic accountability. This is a moderate-weight governance observation.

Claim 6: Grandjean: the US exposure limit, even rebased on the Faroes data, “may not be sufficiently protective”; corrected for nutrient masking and exposure imprecision, it should be lowered by about 50% (p. 122)#

Original claim. The 2000 National Research Council panel endorsed EPA’s 0.1 µg/kg/day reference dose on Faroes data. Grandjean argues that confounding by seafood nutrients and error in measuring exposure both bias the estimated toxicity downward, and that correcting for them implies a limit about 50% lower (Grandjean and Budtz-Jørgensen 2007) (p. 122). The digest notes his declared competing interest as the Faroes principal investigator.

Subsequent developments. - Limits unchanged. - The US EPA oral reference dose for methylmercury remains 1 × 10⁻⁴ mg/kg-day, last updated 27 July 2001. An IRIS reassessment began with an assessment plan (April 2019) and protocol (May 2020), and is “in step 1”, with public comment and peer review “TBD”. (EPA IRIS, methylmercury, accessed 25 Sep 2026) - The US Agency for Toxic Substances and Disease Registry (ATSDR) 2024 minimal risk level is 0.1 µg Hg/kg bw/day, the same as the EPA reference dose. - In a June 2026 statement, EFSA found “no clear indication” that the ATSDR value is more robust than EFSA’s 2012 tolerable weekly intake of 1.3 µg/kg bw, which “remains valid”. EFSA added that new evidence “might merit an update of the 2012 Opinion”. (EFSA statement, 22 Jun 2026) - A 2026 review of US EPA, ATSDR, Health Canada, JECFA and EFSA values finds “good agreement” between them. It notes that most still rest on the same 1980s–90s birth cohorts, and that “several organizations” are reassessing. (Blechinger et al., Arch Toxicol 2026) - No agency has adopted a 50% cut. EFSA’s 2012 value was set shortly before this volume appeared, and is about 20% below the long-standing FAO/WHO Joint Expert Committee on Food Additives (JECFA) value of 1.6 µg/kg bw/week. - Evidence broadly in line with Grandjean: - The Faroes cohort at age 22 showed persisting deficits: about 2.2 IQ points per tenfold increase in prenatal exposure (Debes et al., Cortex 2016). - A 2026 meta-analysis by EPA-affiliated authors of five newer, lower-exposure cohorts estimates −0.407 IQ points per µg/L maternal blood. That is about ten times the 2007 Axelrad slope used in cost-benefit analysis, but the 95% CI (−0.962 to 0.147) includes zero. (Kopylev et al., Expo Health 2026) - A companion meta-analysis finds a statistically significant decrement in language and verbal function (Kopylev & Segal, Environ Health 2025). - Evidence against or complicating: - A 2022 systematic review of 32 prospective studies (to age 5) found the evidence for harm “weak”. It concluded that “dietary mercury exposure during pregnancy is unlikely to be a risk factor for low neurodevelopmental functioning in early childhood”. (Dack et al. 2022) - A 2025 systematic review found seafood during pregnancy associated with better child development (O’Connor et al., Adv Nutr 2025). That supports the “masking” premise but also the case for benefit-risk advice over tighter limits alone.

Verdict: contested. The concern has some new support: persistent effects to age 22, and a steeper but non-significant slope at lower exposures. No regulator has acted on the proposed 50% reduction, and independent systematic reviews find weak evidence of harm at typical dietary exposures. The main reassessments (EPA IRIS, a possible EFSA update) are unfinished.

Implication for weight. Treat Grandjean’s specific number as a respected minority position from an interested party. His underlying methodological points are more durable: - measurement error biases dose-response estimates toward the null; - beneficial co-exposures can mask harm; - demands for more data can serve delay (p. 123).

These carry moderate weight as general mechanisms (digest insight 14). Castaño’s contrasting framing, which calls low-dose effects “assumed” (p. 130), also remains live. The two panels’ implicit disagreement has not been resolved.

Claim 7: Selin: future Asian economic development, especially in China, “could lead to dramatic increase in mercury emissions”; rising gold prices can increase mercury use in artisanal gold mining (p. 127)#

Original claim. Global emissions “have remained relatively constant since 1990”, reflecting Asian increases offsetting declines in North America and Europe. Chinese development “could lead to dramatic increase”; gold prices can increase ASGM mercury use (p. 127).

Subsequent developments. - The premise of flat emissions was challenged. Zhang et al. (2016) found that observed declines in atmospheric mercury (1–2% a year at North American and European sites) were “inconsistent with current global emission inventories indicating flat or increasing emissions”. Their corrected inventory shows a 20% decrease in global emissions from 1990 to 2010, with North American and European declines “offsetting the effect of increasing emissions in Asia”. (PNAS 2016) - China rose, then fell. - China’s emissions rose from 147 t in 1978 to 530 t in 2014 (Wu et al., ES&T 2016). - They then declined “since the 2010s” (Qiu et al. 2025). - The “dramatic increase” from China therefore did not happen after about 2013. - Growth moved elsewhere. Global emissions increased 330% from 1960 to 2021. In 2021, Global South countries excluding China produced two-thirds of emissions while accounting for one-fifth of the world economy. Business-as-usual growth there could raise global emissions 10–50% by 2030. (Qiu et al. 2025) - ASGM remains the largest single source. - The 2018 Global Mercury Assessment put 2015 anthropogenic emissions at 2,220 t, about 20% above an updated 2010 estimate of 1,810 t. - ASGM emitted about 835 t in 2015: 41% from South America, 30% from sub-Saharan Africa and 26% from East and Southeast Asia. - However, the assessment’s experts judged that the apparent 2010–2015 ASGM increase was “largely associated with improved information rather than a significant increase in real world emissions”. (AMAP/UNEP, Technical Background Report for the Global Mercury Assessment 2018, ch. 3, pp. 3-21–3-22 and 3-27) - Gold-price link not tested. I found no study in the sources I could reach that quantifies the relationship between gold prices and ASGM mercury use after 2013. COP-6’s turn to gold supply-chain traceability shows the Parties treat ASGM demand as a policy lever (MOE, 13 Nov 2025).

Verdict: partly held up. - Right: the direction of risk. Growth in developing economies and ASGM became the centre of the problem. - Wrong: the specific China forecast. Its emissions peaked and fell. - Challenged: the baseline premise of flat post-1990 global emissions. - Untested here: the gold-price mechanism.

Implication for weight. The general lesson that problems migrate with industrialisation and informal economic activity holds up, and has moderate weight. National trajectories can reverse quickly when co-benefit policies bite. China’s air-pollution controls reduced mercury as a side effect, much as Selin notes 1980s air-pollution controls did in Europe and North America (p. 127). So forecasts of “dramatic increase” for a single country should be held loosely.

Claim 8: Selin: human disturbance of the mercury cycle is very long-lived (3,000–10,000 years to return to deep-ocean sediments), yet some ecosystems respond quickly to reduced inputs; for example, north-eastern US fish mercury fell between 1999 and 2004 (p. 126)#

Subsequent developments. - Long timescales confirmed. - Tuna mercury has been essentially flat from 1971 to 2022 despite falling emissions and deposition. Modelling attributes this to “the inertia of surface ocean mercury… supplied by legacy mercury that accumulated in the subsurface ocean over centuries”. (Médieu et al. 2024) - Geyman et al. (2025) estimate that anthropogenic enrichment of the ocean was about 270% around 2010. Most mercury released to land and water before 2010 “remains sequestered at contaminated sites” but could be mobilised by climate change. (ES&T 2025) - Angot et al. (2018), co-authored by Selin, show that each five-year delay in peak emissions defers the return to 2010 global deposition by about four more years (ES&T 2018). - Fast local responses confirmed. Hutcheson et al. (2014) extended the cited north-eastern US series to 2011: - In a Massachusetts “hotspot”, largemouth bass mercury fell 44% (in 13 of 16 lakes) and yellow perch 43%. - Point-source emissions there fell 98% from the early 1990s to 2008. - Declines elsewhere in the state were smaller (13–19%). - (ES&T 2014) - New complication. Ocean warming and overfishing-driven diet shifts can raise methylmercury in predators independently of emissions. Examples include up to a 23% rise in Atlantic cod and a modelled 56% rise in bluefin tuna. (Schartup et al., Nature 2019) - The 2026 liquefaction at Minamata Bay (Claim 4) is a local illustration of how legacy mercury stores need permanent management.

Verdict: held up, and reinforced by later evidence on both halves: slow global recovery, fast local response. Climate and fisheries dynamics now complicate any reading of trends as a measure of policy success.

Implication for weight. The two-timescale point carries strong weight: interventions can show quick local results while the global legacy persists for centuries. It explains why the Convention’s effect is hard to see in ocean fish (Claim 5). It also warns that measures of success need to be matched to the timescale of the system concerned.

Claim 9: Castaño: the DEMOCOPHES pilot (17 EU member states) would report by the end of 2012, giving the first harmonised European mapping of mercury in mothers and children to inform the treaty and national strategies (p. 130)#

Subsequent developments. - Reported, a little late. Fieldwork ran 2011–2012 on 1,844 children and their mothers, with results in 2013–2015: - Den Hond et al. (2015) reported that “most biomarker concentrations were below the health-based guidance values”, with wide variability across Europe (EHP 2015). - Castaño et al. (2015) analysed hair mercury in 1,799 mother-child pairs. Exposure correlated strongly with consumption of fish and marine products, with higher levels in southern Europe. Among mothers eating fish only once a week, the 95th percentile was 0.55 µg/g, about half the US EPA-derived 1 µg/g hair benchmark. (Environ Res 2015) - Same data, two readings. Grandjean was a co-author (with Castaño) of an analysis of the DEMOCOPHES data that estimated: - more than 1.8 million EU births a year above 0.58 µg/g hair mercury; - about 200,000 above 2.5 µg/g; - benefits of exposure prevention worth €8–9 billion a year.

(Bellanger et al., Environ Health 2013) The same dataset thus supported both Castaño’s “most below guidance” reading and a precautionary economic case. - Institutionalised. - HBM4EU (2017–2022) produced a methylmercury risk assessment. It found southern Europe close to or above guidance values, but the European population generally below EFSA’s tolerable intake. (Domínguez-Morueco et al. 2022) - HBM4EU also ran a randomised trial of dietary advice to pregnant women in five high-fish-consumption countries (HBM4EU-MOM) (Katsonouri et al. 2023). - National repeated biomonitoring is now presented as evidence for Convention effectiveness evaluation (Usenik et al. 2026).

Verdict: held up. The pilot delivered the promised harmonised European picture, a year or two after the stated date, and led to a standing European biomonitoring capacity.

Implication for weight. Castaño’s claim that better exposure information is a prerequisite for policy was borne out in the sense that the infrastructure was built and used. The data did not settle the normative dispute with Grandjean. Both sides read the same numbers through different thresholds and dose-response assumptions. That supports a technology-neutral point: harmonised measurement narrows factual disputes but does not resolve disagreements about thresholds and values (moderate weight).

Claim 10: Historical claims: 488 tonnes of mercury discharged to the sea 1932–1968; production stopped in May 1968 “for commercial reasons” and only then (26 September 1968) did the government accept causation, 12 years after the cause was identified (pp. 105, 115)#

Subsequent developments (checked against official Japanese sources): - Dates confirmed. - The Kumamoto prefectural chronology records: - June 1966: Chisso’s acetaldehyde wastewater recirculation system completed; - 18 May 1968: Chisso stops acetaldehyde production; - 26 September 1968: the health ministry announces that the cause was the Minamata factory’s effluent.

([Kumamoto chronology](https://www.pref.kumamoto.jp/uploaded/life/2711_846954_misc.pdf))

• The Minamata City Municipal Museum notes that in May 1968 Japan’s last mercury-catalysed acetaldehyde plants, Chisso Minamata and Denki Kagaku’s Omi plant, had stopped operating, and describes the official view as coming in “the 12th year” after May 1956 (Museum booklet, Jan 2023).

This supports the chapter’s point that acknowledgement followed obsolescence of the process. Neither source states the commercial motive, which the chapter takes from Arima (1979). - The amount is not corroborated. I could not access the chapter’s source (Miyazawa 1996). The museum booklet says the mercury “deposited” in Minamata Bay from 1932 to 1968 “is said to be about 70–150 tonnes, or more”. That is a different quantity (sediment in the bay, not total discharge to the Shiranui Sea), so the two are not directly contradictory. But 488 t is at the high end and appears single-sourced in the chapter. - When discharges ended is disputed. MOE’s official English history (2002) states that after the recirculation system was completed in 1966, “water effluent containing methylmercury compound had not been discharged outside of the plant in principle” (MOE, Minamata Disease: The History and Measures, ch. 4). The chapter’s framing of discharge continuing to 1968 therefore conflicts with the government’s account. I could not independently resolve which is right.

Verdict: partly held up. - Confirmed: the key dates and the “12 years” interval. - Consistent with official sources: that acknowledgement came only after the process was obsolete. - Not corroborated: the 488-tonne figure. - Disputed by the government: the end date of discharges.

Implication for weight. The “acknowledgement awaits obsolescence” mechanism (digest insight 7) rests on the timing, and the timing is confirmed; it keeps moderate weight. Avoid quoting 488 t as settled. If a figure is needed, give the range and say what each estimate measures.

Summary of verdicts#

Technology-neutral lessons this check supports (for later use as a lens)#

1. Once fixed, a case definition becomes an institution. - What happened: the 1977 criteria survived a professional-society verdict (1998), a Supreme Court ruling that they are non-exclusive (2013), a lower court ordering certification (2026) and the delegated administrator’s requests to review them. - What sustained them: the central authority’s appeal rationale was “uniform operation”, not medical validity. - References: pp. 94, 109–110; Claim 2.
2. Relief without recognition buys time, not closure. - Two “final solutions” (1995, 2009) each paid tens of thousands without admitting liability. - Each was followed by new litigation and demands for a third settlement. - References: pp. 109–110; Claims 1 and 3.
3. A statutory duty to investigate can be discharged in form while being deferred in substance. - Seventeen years passed between the 2009 mandate and a 32-person feasibility pilot. - Closed expert panels chose the methods. - The survey’s purpose was framed as reassurance, and its exclusion criteria screen out many older residents. - References: pp. 92, 110; Claim 4.
4. Counting choices shape perceived scale. - The count was passive and application-based, adjudicated by the payer. - For decades official reports showed only living patients. - Relief populations were kept off the patient register. - Together these make a large harm look small and finished. - References: pp. 94, 107–110; Claim 1.
5. Paying for the harm and ruling on it are still entangled. - The state co-finances the polluter’s liabilities. Under a 2000 Cabinet understanding, Chisso pays patient compensation out of ordinary profit first, and repays Kumamoto Prefecture’s loans only “within the possible range”. In May 2026 a reporter noted that the formula implied full deferral that year. (MOE press conference, 15 May 2026) - The state also defends the certification criteria and appeals the rulings against them. - References: pp. 108–110.
6. Procedural respect is a recurring failure point, and a recoverable one. - The failures: the scripted microphone cut-off (May 2024) and disputed remarks by officials (May 2026). - The recovery: apology, longer dialogues and indexed allowances. - Recovery did not reach the substantive dispute. - References: pp. 92, 114 (“democratic deficit”); Claims 1, 3 and 4.
7. Legacy contamination sets its own timescale. - Global recovery takes centuries because of ocean inertia. - Local recovery can take years, as when fish mercury fell after point sources closed. - Contained stores need permanent vigilance, as the 2026 liquefaction showed. - Success metrics must match the system’s response time. - References: p. 126; Claims 5 and 8.
8. Harmonised measurement narrows factual disputes but not normative ones. - Castaño and Grandjean read the same exposure data through different thresholds and dose-response assumptions. - References: pp. 122, 130; Claims 6 and 9.
9. Forecasts of where a problem moves hold up better than forecasts of which country drives it. - Emissions migrated to informal and developing-economy sources, as forecast. - China, the named driver, reversed course through co-benefit controls. - References: p. 127; Claim 7.

Sources#

Japanese government, courts and official bodies - Kumamoto Prefecture, Status of certification work (as of end August 2026), PDF: https://www.pref.kumamoto.jp/uploaded/attachment/318159.pdf (page https://www.pref.kumamoto.jp/soshiki/48/50008.html, updated 10 Sep 2026) - Kumamoto Prefecture, Chronology of the Minamata disease problem (1956–FY2025): https://www.pref.kumamoto.jp/uploaded/life/2711_846954_misc.pdf (page https://www.pref.kumamoto.jp/soshiki/47/1711.html) - Minamata City Municipal Minamata Disease Museum, Minamata Disease: Its History and Lessons 2022 (Jan 2023): https://www.city.minamata.lg.jp/mdmm/kiji0034102/3_4102_21368_up_ntmgqo27.pdf - Niigata Prefecture, press release on certification decisions (10 Aug 2026): https://www.pref.niigata.lg.jp/uploaded/attachment/506651.pdf - Niigata Prefecture, governor’s comment on the Niigata District Court ruling (12 Mar 2026): https://www.pref.niigata.lg.jp/uploaded/attachment/485001.pdf - Niigata Prefecture, governor’s comment on filing an appeal (26 Mar 2026): https://www.pref.niigata.lg.jp/uploaded/attachment/487770.pdf - Ministry of the Environment (MOE), Minamata disease policy page: https://www.env.go.jp/chemi/minamata.html (accessed 25 Sep 2026) - MOE, Results of determinations under the Special Measures Act (determinations ended Jan 2018): https://www.env.go.jp/content/000161039.pdf - MOE, Act on Special Measures for the Relief of Minamata Disease Victims (2009), text: https://www.env.go.jp/content/900410653.pdf - MOE, Relief measures policy (Cabinet decision 16 Apr 2010): https://www.env.go.jp/content/900410656.pdf - MOE, Feasibility study of an epidemiological survey on methylmercury health effects (1 Oct 2025): https://www.env.go.jp/content/000368969.pdf - MOE, Feasibility study: preliminary results (Feb 2026): https://www.env.go.jp/content/000381493.pdf - MOE, MEG/MRI health-survey research group, FY2024 report summary: https://www.env.go.jp/content/000308991.pdf - MOE, press release on revision of medical allowances (21 Apr 2026): https://www.env.go.jp/press/press_04454.html - MOE, minister’s message on allowance revision (26 Dec 2025): https://www.env.go.jp/content/000366661.pdf - MOE, press release on COP-6 results (13 Nov 2025): https://www.env.go.jp/press/press_01704.html - MOE, press release on OESG-2 in Minamata (11 Mar 2025): https://www.env.go.jp/press/press_04593.html - MOE, overview of the Minamata Convention: https://www.env.go.jp/chemi/tmms/convention.html - MOE, Minamata Disease: The History and Measures (2002), ch. 4: https://www.env.go.jp/en/chemi/hs/minamata2002/ch4.html - MOE, minister’s press conference transcripts: - 4 Oct 2024: https://www.env.go.jp/annai/kaiken/kaiken_00248.html - 28 Jan 2025: https://www.env.go.jp/annai/kaiken/kaiken_00279.html - 9 May 2025: https://www.env.go.jp/annai/kaiken/kaiken_00307.html - 3 Oct 2025: https://www.env.go.jp/annai/kaiken/kaiken_00329.html - 25 Nov 2025: https://www.env.go.jp/annai/kaiken/kaiken_00344.html - 13 Mar 2026: https://www.env.go.jp/annai/kaiken/kaiken_00370.html - 24 Mar 2026: https://www.env.go.jp/annai/kaiken/kaiken_00373.html - 28 Apr 2026: https://www.env.go.jp/annai/kaiken/kaiken_00382.html - 15 May 2026: https://www.env.go.jp/annai/kaiken/kaiken_00385.html - 5 Jun 2026: https://www.env.go.jp/annai/kaiken/kaiken_00392.html - 12 Jun 2026: https://www.env.go.jp/annai/kaiken/kaiken_00394.html - 10 Jul 2026: https://www.env.go.jp/annai/kaiken/kaiken_00404.html - Supreme Court of Japan, judgment of 16 Apr 2013, Heisei 24 (Gyo-Hi) No. 245: - summary: https://www.courts.go.jp/hanrei/83193/detail2/index.html - full text: https://www.courts.go.jp/assets/hanrei/hanrei-pdf-83193.pdf - Supreme Court of Japan, judgment of 16 Apr 2013, Heisei 24 (Gyo-Hi) No. 202: https://www.courts.go.jp/hanrei/83192/detail2/index.html

International - UN Treaty Collection, Minamata Convention on Mercury, status as at 26 Sep 2026: https://treaties.un.org/Pages/ViewDetails.aspx?src=TREATY&mtdsg_no=XXVII-17&chapter=27&clang=_en - AMAP/UN Environment (2019), Technical Background Report for the Global Mercury Assessment 2018: https://www.amap.no/documents/doc/technical-background-report-for-the-global-mercury-assessment-2018/1815 - US EPA IRIS, methylmercury (reference dose last updated 27 Jul 2001; reassessment status), accessed 25 Sep 2026: https://iris.epa.gov/ChemicalLanding/&substance_nmbr=73 - EFSA (2026), Statement on the findings presented in the Toxicological Profile for Mercury issued by the US ATSDR, EFSA Journal, 22 Jun 2026: https://doi.org/10.2903/j.efsa.2026.10156

Peer-reviewed literature - Angot H et al. (2018) Global and local impacts of delayed mercury mitigation efforts. ES&T: https://doi.org/10.1021/acs.est.8b04542 - Bellanger M et al. (2013) Economic benefits of methylmercury exposure control in Europe. Environ Health 12:3: https://doi.org/10.1186/1476-069X-12-3 - Blechinger SR et al. (2026) Neurodevelopmental effects of methylmercury: review of PoDs and TRVs. Arch Toxicol, 10 Mar 2026: https://doi.org/10.1007/s00204-026-04345-8 - Castaño A et al. (2015) Fish consumption patterns and hair mercury levels in children and their mothers in 17 EU countries. Environ Res: https://doi.org/10.1016/j.envres.2014.10.029 - Dack K et al. (2022) Prenatal mercury exposure and neurodevelopment up to the age of 5 years: a systematic review. IJERPH: https://doi.org/10.3390/ijerph19041976 - Debes F, Weihe P, Grandjean P (2016) Cognitive deficits at age 22 years associated with prenatal exposure to methylmercury. Cortex: https://doi.org/10.1016/j.cortex.2015.05.017 - Den Hond E et al. (2015) First steps toward harmonized human biomonitoring in Europe (DEMOCOPHES). EHP: https://doi.org/10.1289/ehp.1408616 - Domínguez-Morueco N et al. (2022) Methylmercury risk assessment based on European human biomonitoring data. Toxics: https://doi.org/10.3390/toxics10080427 - Futatsuka M, Eto K, Uchino M (2025) Discussion about “Chronic Feature of Minamata Disease”. Nihon Eiseigaku Zasshi: https://doi.org/10.1265/jjh.24013 - Geyman BM et al. (2025) Cumulative anthropogenic impacts of past and future emissions and releases on the global mercury cycle. ES&T, 22 Apr 2025: https://doi.org/10.1021/acs.est.4c13434 - Hutcheson MS et al. (2014) Temporal and spatial trends in freshwater fish tissue mercury concentrations associated with mercury emissions reductions. ES&T: https://doi.org/10.1021/es404302m - Katsonouri A et al. (2023) HBM4EU-MOM study design. Int J Hyg Environ Health: https://doi.org/10.1016/j.ijheh.2023.114213 - Kopylev L, Segal D (2025) Prenatal methylmercury exposure and language/verbal function: a meta-analysis. Environ Health, 29 Sep 2025: https://doi.org/10.1186/s12940-025-01228-w - Kopylev L et al. (2026) Relationship between prenatal methylmercury exposure and IQ: a meta-analysis. Exposure and Health, 21 Apr 2026: https://doi.org/10.1007/s12403-026-00781-0 - Médieu A et al. (2024) Stable tuna mercury concentrations since 1971 illustrate marine inertia and the need for strong emission reductions under the Minamata Convention. ES&T Letters, 21 Feb 2024: https://doi.org/10.1021/acs.estlett.3c00949 - O’Connor LE et al. (2025) Seafood during pregnancy and lactation and child neurocognitive development: a systematic review. Adv Nutr: https://doi.org/10.1016/j.advnut.2025.100414 - Qiu X et al. (2025) Declines in anthropogenic mercury emissions in the Global North and China offset by the Global South. Nat Commun, 30 Jan 2025: https://doi.org/10.1038/s41467-025-56274-2 - Schartup AT et al. (2019) Climate change and overfishing increase neurotoxicant in marine predators. Nature: https://doi.org/10.1038/s41586-019-1468-9 - Takaoka S (2024) Responses to “A Review of the Book ‘Minamata Disease and the Responsibility of Medical Authorities’”. Nihon Eiseigaku Zasshi: https://doi.org/10.1265/jjh.24005 - Takaoka S (2025) On the description of “Forum” by Drs Futatsuka, Eto and Uchino. Nihon Eiseigaku Zasshi: https://doi.org/10.1265/jjh.25002 - Takaoka S et al. (2024) Health effects of methylmercury in coastal areas of the Yatsushiro Sea, far from Minamata. Toxics: https://doi.org/10.3390/toxics12100751 - Usenik V et al. (2026) Human biomonitoring in support of the Minamata Convention: a case of phasing out dental amalgam. Environ Health, 5 Jan 2026: https://doi.org/10.1186/s12940-025-01255-7 - Wu Q et al. (2016) Temporal trend and spatial distribution of speciated atmospheric mercury emissions in China during 1978–2014. ES&T: https://doi.org/10.1021/acs.est.6b04308 - Yorifuji T, Takaoka S, Grandjean P (2018) Accelerated functional losses in ageing congenital Minamata disease patients. Neurotoxicol Teratol: https://doi.org/10.1016/j.ntt.2018.08.001 - Yorifuji T et al. (2025) Prenatal methylmercury poisoning and neurocognitive impairment in Minamata. Sci Total Environ, 28 May 2025: https://doi.org/10.1016/j.scitotenv.2025.179743 - Yorifuji T et al. (2026) Long-term neurological and neurocognitive deficits in adults prenatally exposed to methylmercury: Minamata disease. Neurotoxicol Teratol, 27 Feb 2026: https://doi.org/10.1016/j.ntt.2026.107590 - Zhang Y et al. (2016) Observed decrease in atmospheric mercury explained by global decline in anthropogenic emissions. PNAS: https://doi.org/10.1073/pnas.1516312113

News and secondary sources (used where no primary source was reachable) - Japan Times tag page, Minamata disease (headlines and leads for 27 Sep 2023, 22 Mar 2024, 19 Apr 2024 and 1 May 2025; full texts expired): https://www.japantimes.co.jp/tag/minamata-disease/ - Japan Times editorial, “Wider Minamata relief needed” (22 Apr 2013): https://www.japantimes.co.jp/opinion/2013/04/22/editorials/wider-minamata-relief-needed/ - Japan Times editorial, “New burden for Minamata victims” (17 Mar 2014): https://www.japantimes.co.jp/opinion/2014/03/17/editorials/new-burden-for-minamata-victims/ - Japan Times editorial, “Relief for Minamata victims” (21 Jan 2017): https://www.japantimes.co.jp/opinion/2017/01/21/editorials/relief-minamata-victims/ - Japan Times, “Minamata mercury treaty finds skeptics” (7 Oct 2013): https://www.japantimes.co.jp/news/2013/10/07/national/minamata-mercury-treaty-finds-skeptics/ - Japan Times, “Mercury still threat, Abe assurances or not” (21 Oct 2013): https://www.japantimes.co.jp/news/2013/10/21/reference/mercury-still-threat-abe-assurances-or-not/ - Japan Times, “Japan to begin Minamata disease survey as early as this year” (30 Apr 2026): https://www.japantimes.co.jp/news/2026/04/30/japan/japan-minamata-disease-survey/ - Japan Times, “Decades later, the dispute over Minamata disease remains unresolved” (1 May 2026): https://www.japantimes.co.jp/news/2026/05/01/japan/minamata-disease-70-years/ - Asahi Shimbun, Kumamoto District Court ruling (22 Mar 2024): https://www.asahi.com/articles/ASS3Q26DDS3KTLVB005.html - Asahi Shimbun, MOE health survey and the Kumamoto earthquake (15 Sep 2026): https://www.asahi.com/articles/ASV9H3FCFV9HTIPE001M.html - Asahi Shimbun, liquefaction of Minamata Bay reclaimed land (16 Sep 2026): https://www.asahi.com/articles/ASV9H3CMJV9HTIPE018M.html - Minamata Disease Center Soshisha, list of lawsuits (updated June 2022, entries to 2024): https://www.soshisha.org/jp/about_md/suits