Hindsight check: LL2-06 (Ch 6 Beryllium’s ‘public relations problem’)#

Source section: EEA, Late lessons from early warnings: science, precaution, innovation (EEA Report No 1/2013), Ch 6. The chapter is by David Michaels and Celeste Monforton (pp. 131–144), followed by a panel by Tee L. Guidotti (pp. 145–150). Check window: publication (2013) to late September 2026. Checked: 25 September 2026.

Method note. - General web search was unavailable for this pass because the session’s search budget was exhausted. I retrieved sources by fetching primary repositories directly: - the Federal Register (its API, plus govinfo.gov full texts); - EUR-Lex; - the EU Advisory Committee on Safety and Health at Work (ACSH) document repository; - US Geological Survey (USGS) Mineral Commodity Summaries; - SEC EDGAR (Materion’s annual 10-K reports); - reginfo.gov (the Unified Agenda); - the US Code (uscode.house.gov); - the regulations.gov API (metadata only); - Europe PMC and Crossref, for peer-reviewed abstracts and funding metadata. - Some sources refused automated retrieval: - OSHA news releases (403) and the regulations.gov PDF of the 2012 Materion–USW letter (403). I use the Federal Register’s description of that letter instead. - The Department of Labor’s EEOICPA compensation statistics pages (403/404). - ECHA (403). - The Chest journal site (403). For two Chest letters I cite only the title and authors. - EUR-Lex refused later requests (HTTP 202, empty), so I could not confirm the consolidated Carcinogens Directive. Where a point rests on an abstract or on metadata only, I say so. - Annex 3 does not apply. Beryllium is a new case in the 2013 volume, so there is no Annex 3 update. Annex 3 covers only the 2001 cases. - I read only the section digest, the section’s source extract and web sources. Page numbers are report pages of the 2013 volume. - Conflict of roles. Michaels, the chapter’s lead author, was head of OSHA when the chapter was published. He signed the 2017 OSHA rule, on 14 December 2016 [S1]. OSHA’s later findings therefore partly vindicate a position the author himself acted on. I flag below where support comes from bodies independent of him: the EU scientific committee SCOEL, the DOE, NIOSH, the American Thoracic Society and academic groups.

Overview#

The chapter makes three kinds of claim. Thirteen years on, they have fared differently.

1. The empirical core has held and in places strengthened. - 2 μg/m³ did not protect workers (pp. 138–140). - In January 2017 OSHA found “significant risk” at the old limit. It cut the 8-hour limit tenfold to 0.2 μg/m³, added a 15-minute limit (STEL) of 2.0 μg/m³ and an action level of 0.1 μg/m³, and extended coverage to construction and shipyards [S1]. - The rule survived a 2017 freeze and a proposal to strip its extra protections for construction and shipyards. It was amended in 2018 and 2020 without weakening the limits [S2–S8]. - The EU set a binding 0.2 μg/m³ limit (inhalable fraction) in 2019, with a transitional 0.6 μg/m³ until 11 July 2026 [S17]. - No safe level identified (p. 139). OSHA itself judged that significant risk remains at 0.2 μg/m³ and stopped there only because it could not show a lower limit was technologically feasible [S1]. The EU’s scientific committee recommended a limit ten times lower still, 0.02 μg/m³ [S19]. Among the EU’s advisory social partners, the workers’ group called 0.2 μg/m³ “not protective against Beryllium sensitisation” [S19]. - Downstream and life-cycle exposures (pp. 132, 139–140). These were written into the US rule’s scope: foundries, secondary smelting, dental laboratories, aluminium production, coal-fired power and abrasive blasting [S1]. New unexpected sources have since appeared: beryllium-rich concrete dust in Germany [S36], poor dust control in formal e-waste recycling in Canada [S38], and construction trades at nuclear sites [S35].

2. The policy prescription has not been followed and has lost ground. - End industrial use except where substitution is impossible (pp. 139–140). - US apparent consumption was about 220 t in 2012 and about 230 t in 2025 [S21, S22]. - Beryllium is on both the US 2025 critical-minerals list and the EU 2023 critical raw materials list [S23, S20]. - US defence funding (about 70% of a roughly $105 million plant) re-established domestic metal production [S24]. - The producer now books a federal production tax credit for critical minerals [S26]. - USGS’s substitution text is essentially unchanged from 2013 to 2026: alternatives “can result in substantially reduced performance” [S21, S22]. - The mainstream route has been tighter control, not elimination.

3. The contested interpretive claims are only partly resolved. - Discount the science of financially interested parties (p. 140) versus Guidotti’s call to audit it (p. 148). Later events support both. - For Michaels: industry-linked epidemiology kept contesting lung-cancer risk after 2013, some of it by Exponent staff with Materion co-authors or funding [S45–S47]. NIOSH rebutted it [S50]. - For Guidotti: - OSHA’s rule was “based, in part” on a February 2012 draft standard written jointly by Materion and the United Steelworkers (USW) [S1]. - NIOSH–industry studies produced protective findings, notably that sensitisation also occurs through skin contact [S27, S31]. - OSHA weighed industry studies on their methods rather than their funding [S1]. - Companies need exit routes (pp. 149–150). This fits the sequence only loosely. - By December 2011 Materion reported no pending beryllium lawsuits [S24]. - A federal compensation statute already required claimants to give up tort suits against beryllium vendors [S16]. - The company’s turn to support a tenfold-lower limit came after that, but no formal legacy-liability deal accompanied it. The standard also codified what Materion said were its existing practices [S25]. - Standards become “sticky” (p. 150). This is well supported: - the 2 μg/m³ number was enforced by OSHA until 2018; - DOE’s 1999 rule is still unrevised after proposals in 2016 and 2023 [S13–S15]; - the EU chose 0.2 over its scientific committee’s 0.02, and lengthened the Commission’s proposed five-year transition to seven [S17–S19]. - Continuous improvement was not adopted anywhere I found. - Short peaks drive risk (p. 134). This remains plausible but unproven. - Peak metrics are associated with sensitisation, but they cannot be separated from skin exposure [S31]. - OSHA said the literature does not address STEL-level differences [S1]. - SCOEL proposed a 0.2 μg/m³ STEL; the EU set none [S19, S17].

Implication for weight. Give strong weight to the chapter’s diagnostic lessons: provisional limits harden into anchors, detection limits breed false reassurance, and harm migrates downstream. Later regulatory records independent of the authors confirm these. Treat the recommendation to end most use as advocacy that policy has moved away from, because supply security now dominates. Treat the chapter’s second lesson (discount interested parties) as half right. Interested-party science did remain partisan on contested endpoints. But the most consequential later step, a tenfold-lower limit, was co-drafted by the producer and a union. It was then vetted by a regulator that judged studies on their merits, which is closer to Guidotti’s auditing model. Realised health benefits cannot yet be checked: OSHA’s projected 90 deaths and 46 illnesses avoided a year apply “once it is fully effective” [S1], and no retrospective evaluation exists.

Claim-by-claim#

Claim 1: OSHA’s 2 μg/m³ limit did not adequately protect workers from chronic beryllium disease (CBD), and OSHA had yet to propose a new standard as of 2012 (pp. 138–140, Table 6.1)#

Original claim (pp. 138–140). - In 1998 OSHA wrote that its 2 μg/m³ limit “does not adequately protect” workers (p. 138). Researchers including some affiliated with Brush Wellman published further evidence (p. 139). A Brush-supported 2006 review conceded “insufficient protection” (p. 139). - Table 6.1 closes with “2012 OSHA has yet to propose new workplace beryllium standard” (p. 140). Guidotti notes “at least 35 years of deliberation” (p. 145).

Subsequent developments - Proposal and final rule. - OSHA proposed a new standard on 7 August 2015 (80 FR 47566) [S10]. It published the final rule on 9 January 2017 (82 FR 2470) [S1], with: - an 8-hour time-weighted average (TWA) permissible exposure limit (PEL) of 0.2 μg/m³; - a 15-minute STEL of 2.0 μg/m³; - an action level of 0.1 μg/m³; - ancillary provisions: exposure assessment, respiratory and dermal protection, housekeeping, medical surveillance including the beryllium lymphocyte proliferation test (BeLPT), medical removal, and hazard communication. - It covers general industry, construction and shipyards. OSHA estimates about 62,000 exposed workers in about 7,300 establishments, including about 12,000 in construction and shipyards [S9]. - Finding on the old limit. OSHA found “significant risk of beryllium sensitization, CBD, and lung cancer from a 45-year (working life) exposure” at the old 2 μg/m³ limit [S1]. Its lung-cancer modelling, drawn from NIOSH’s Schubauer-Berigan et al. (2011), estimated 33–170 excess lung cancers per 1,000 workers at 2 μg/m³ [S1]. - Delays and attempted rollback, 2017–2020. - Effective dates slipped under the 2017 regulatory freeze: first to 21 March, then to 20 May 2017 [S2, S3]. - On 27 June 2017 OSHA proposed revoking the ancillary provisions for construction and shipyards while keeping the new limits [S4]. On 30 September 2019 it declined to revoke them, finding “not complete overlap” with other standards [S6, S8]. - Enforcement came in stages: - the PEL and core provisions from 11 May 2018; - most other provisions from 12 December 2018; - showers and change rooms from 11 March 2019; - engineering controls from 10 March 2020 [S7]. - The 2018 and 2020 amendments clarified trace-material and dermal provisions. OSHA said they would “maintain or enhance worker protections overall” [S5, S7, S8]. - No later change to the limits. A Federal Register API search of all agencies (June 2024 to September 2026) found no OSHA rulemaking to revise the beryllium limits, only routine paperwork renewals [S11]. - The producer’s position. Materion’s 2017 annual report described the rule as having “lowered the permissible exposure by a factor of ten”. It said Materion “was a participant in the development of the new standards, which fundamentally represent our current health and safety operating practices” [S25]. - The EU. It set its first binding beryllium limit in 2019 at 0.2 μg/m³ (inhalable fraction), with a dermal and respiratory sensitisation notation [S17]. National limits had lagged. A survey of five French plants published in 2019 still measured against a French limit of 2 μg/m³, which only 2% of samples exceeded [S52].

Complications - The 2017 rule was issued under the chapter’s own lead author [S1], so it is not independent confirmation. But: - OSHA’s external peer reviewers supported its conclusions; - NIOSH, National Jewish Health, the American Thoracic Society and the dominant producer all backed a lower limit in the record [S1]; - the EU reached the same number through its own scientific and tripartite process [S17–S19].

Verdict: strengthened.

Implications for weight. The factual core of the case (a provisional number persisting decades after its inadequacy was documented) is now confirmed by the regulator’s own formal findings. In the US the 2 μg/m³ figure governed from 1949 (Atomic Energy Commission) and 1971 (OSHA) until enforcement changed in 2018. The lesson is reliable. Its most vivid support comes from an agency the author led, but independent bodies converged on the same conclusion.

Claim 2: No safe level of beryllium exposure can be identified, yet cutting exposure to the lowest achievable levels has reduced sensitisation and CBD (p. 139)#

Original claim (p. 139). - Citing the National Research Council (2008) and Kreiss et al. (2007), the authors write that “it is not possible to estimate a chronic inhalation-exposure level that is likely to prevent BeS and CBD”. BeS is beryllium sensitisation. - Interventions that cut exposure to the lowest achievable levels “have successfully decreased BeS and CBD incidence”. - They note the ACGIH’s 2009 recommended limit (Threshold Limit Value, TLV) of 0.05 μg/m³.

Subsequent developments: residual risk at 0.2 μg/m³ - OSHA (2017). OSHA wrote that the agency “considers the level of risk remaining at the new TWA PEL to still be significant”. It did not go lower “because the Agency could not demonstrate technological feasibility of a lower TWA PEL” [S1]. - Its peer reviewers agreed that “substantial risk of sensitization and CBD were observed in facilities where the highest exposure generating processes had median full-shift exposures around 0.2 μg/m³ or higher”. They also agreed “the greatest reduction in risk was achieved when exposures for all processes were lowered to 0.1 μg/m³ or below” [S1]. - Excess lung cancer at 0.2 μg/m³ was still estimated at 3–30 per 1,000 workers [S1]. - OSHA set medical surveillance at the action level because “significant risk remain[s] at the action level and PEL” [S1]. - SCOEL (EU scientific committee), Recommendation 175 (8 February 2017). It recommended an 8-hour limit of 0.02 μg/m³ (inhalable), ten times lower than the eventual EU and US numbers, and a 15-minute STEL of 0.2 μg/m³ [S19]. - EU ACSH (the tripartite advisory committee), opinion of 31 May 2017 [S19]: - The workers’ group stated that the agreed 0.2 μg/m³ “is not protective against Beryllium sensitisation” and asked for later review towards 0.02 μg/m³. - The employers’ group stated that 0.2 μg/m³ “is protective against chronic beryllium disease”. It also claimed recent studies identify a no-observed-adverse-effect level (NOAEL) “at a higher level”. - DOE registry (sites under the 1999 0.2 μg/m³ trigger, but with legacy exposure mixed in). Through 2013, 407 of 21,453 tested workers (1.9%) were sensitised without CBD and 146 (0.7%) had CBD, with wide variation between sites [S13]. - DOE construction trades. From 1998 to 2020, 262 of 21,854 screened workers (1.20%) were sensitised [S35].

Subsequent developments: does lowering exposure work? - Supporting evidence. - In a NIOSH-authored study with a Materion co-author, a producer’s comprehensive programme cut sensitisation among new hires. The programme, begun in 2000, combined enclosure and ventilation, respiratory and dermal protection, control of dust migration, housekeeping and training. Cross-sectional prevalence was 0.7% (2/298), 2.3% when surveillance data were added, and 0.6% (1/175) among later hires (Thomas et al. 2013) [S27]. - OSHA’s reading of four plants was the same: comprehensive programmes “sharply curtailed new cases of sensitization among newly-hired workers”. By contrast, engineering controls alone that brought median exposures to about 0.2 μg/m³, “with no corresponding emphasis on PPE” (personal protective equipment), “were less effective” [S1]. - Complicating evidence. - In a multi-site case-control study of 444 sensitised workers, 449 with CBD and 890 controls, higher reconstructed exposure was associated with lower odds of sensitisation. CBD was not associated with exposure against controls, although the share of CBD relative to BeS rose with exposure. The authors conclude “reducing exposure may not prevent BeS, [but] it may reduce CBD” (Crooks et al. 2022) [S34]. - A high-exposure cohort showed no exposure-response for BeS or CBD (Widyaningsih et al. 2020) [S33]. - Genetic susceptibility is now firmly established. Some HLA-DPB1 genotypes carrying the E69 variant show CBD prevalence above 40% among exposed workers (Kreiss et al. 2016) [S30]. The American Thoracic Society’s 2014 statement lists higher exposure and the E69 variant as risk factors for progression [S28].

Verdict: strengthened on “no safe level identified” and residual risk at the new limits. It is nuanced on mechanism: the reductions came from programmes that also blocked skin contact, not from lower airborne concentration alone.

Implications for weight. The “no identifiable threshold, so minimise exposure” position is now the regulators’ own. OSHA and the EU’s scientific committee both acknowledge residual risk at 0.2 μg/m³. The lesson should carry the refinement the chapter lacked. What worked was controlling several exposure routes. Airborne concentration, the metric the standard regulates, is an imperfect proxy for the dose that sensitises.

Claim 3: “A more protective standard will help prevent CBD and save lives” (p. 140)#

Original claim (p. 140). “Extensive research has subsequently confirmed the inadequacy of the OSHA standard; a more protective standard will help prevent CBD and save lives.”

Subsequent developments - OSHA’s projection. The final rule is “estimated to prevent 90 fatalities and 46 beryllium-related illnesses annually once it is fully effective”. The 90 deaths are 4 from lung cancer and 86 from CBD. Costs were estimated at $74 million a year and monetised benefits at $561 million a year, for net benefits of $487 million (3% discount rate, 2015 dollars) [S1]. The phrase “once it is fully effective” matters: the benefits accrue over a working lifetime of reduced exposure. - No retrospective evaluation. I found none: no OSHA lookback, no NIOSH or academic study of CBD incidence or mortality before and after the rule (Europe PMC searches, 2017–2026). - Why outcomes are hard to measure. - Death certificates are too sparse. NIOSH counted pneumoconiosis deaths from “other inorganic dust”, the category that includes berylliosis. There were 12 in 1999 and 25 in 2018, with “no evidence of a change in death rates” and too few to analyse berylliosis separately (Bell & Mazurek 2020) [S44]. These counts are far below OSHA’s modelled 86 CBD deaths avoided a year, which points to under-recording, a modelling gap or both. - Misdiagnosis. CBD is often recorded as sarcoidosis, a lung disease with no known cause (NAS 2008, cited in [S1]). - National Jewish Health argued in 2024 that sarcoidosis diagnosed in military personnel may be CBD [S40, title only]. - A 2025 case-case study found beryllium-exposed “sarcoidosis” cases differed from CBD. Only 53% carried the E69 variant, against 92% of CBD cases, so some may be a distinct disease (Mayer et al. 2025) [S39]. - Disease severity. CBD does raise mortality. In a Michigan cohort of 354 people with CBD and 290 with BeS, CBD raised respiratory mortality but BeS did not (TenHarmsel et al. 2024; one co-author from Materion) [S43]. So a reduction in CBD should save lives, if it occurs. - Clinical reviews. A 2020 review said the 2017 rule “may ultimately enhance worker protection” [S41]. A 2026 review states that “despite exposure regulations, new cases continue to emerge across both traditional and nontraditional occupational settings” (Morgan & MacMurdo 2026) [S42]. This fits with the rule not yet being fully effective. It does not measure the rule’s effect.

Verdict: unclear. The claim is plausible and consistent with intervention studies [S27], but it is not yet tested against outcomes.

Implications for weight. Treat the claim as a reasonable expectation, not a demonstrated result. The case also shows a general limit: when the harm has a long latency and is often misclassified, benefits can be modelled but not quickly verified. Surveillance built into the rule is the only route to evidence. The rule’s medical surveillance (BeLPT screening at the action level) could provide that evidence if its data were pooled, but no pooled analysis is published.

Claim 4: Substitute a less toxic material wherever possible; “end industrial use of beryllium, except in circumstances where substitution is impossible”; otherwise cut exposure to the lowest technically feasible level (pp. 139–140)#

Original claim (pp. 139–140). The authors propose, as “prudent public health policy”, that manufacturers substitute a less toxic material “whenever possible” (p. 139). Because secondary users and recyclers are unlikely to manage the risk, they propose ending industrial use except where substitution is impossible (p. 140). The digest classes this as advocacy rather than a lesson drawn from evidence.

Subsequent developments - Consumption did not fall. - US apparent consumption was 220 t in 2012 [S21] and 230 t in 2025, valued at about $360 million [S22]. - Estimated world mine production was about 230 t in 2012, with a USGS caveat that China may have produced more than 60 t [S21]. It was about 430 t in 2025 [S22]. This rise partly reflects USGS adding Brazil and Nigeria and revising figures, so it is not a clean trend. - Uses stayed broad. In 2025 the main end uses were consumer electronics (29%), aerospace and defence (24%), industrial components (17%), automotive electronics (9%) and energy (8%) [S22]. The 2012 profile was similarly broad [S21]. - Substitution unchanged. USGS’s “Substitutes” text is nearly identical in 2013 and 2026. Alternatives exist for many uses (composites, high-strength aluminium, titanium, other copper alloys, aluminium or boron nitride), but some “can result in substantially reduced performance” [S21, S22]. - The state moved to secure supply. - The US Department of Defense co-funded a primary beryllium plant under the Defense Production Act, Title III. The total cost was about $104.9 million, with the DoD providing about 70%; the plant entered service in 2012 [S24]. The 2025 10-K still carries $63.5 million of DoD reimbursement on the balance sheet [S26]. - Beryllium is on the US Final 2025 List of Critical Minerals [S23] and the EU’s 2023 list of critical raw materials [S20]. - Materion reports that it is eligible for the Advanced Manufacturing Production Credit (Inflation Reduction Act of 2022) for critical minerals produced and sold in the US. It books the credit as a reduction in cost of goods sold [S26]. - Some market pressure towards substitution remains. Materion’s 2025 10-K lists substitution as a risk. Customers may choose substitutes because of “health and safety concerns” and “the risk of litigation”. The company adds: “despite numerous studies affirming the safety of beryllium in these products” [S26]. - The regulatory path taken was control. Neither OSHA nor the EU restricted uses. Both set lower exposure limits with ancillary controls [S1, S17]. I found no EU restriction of beryllium uses under chemicals law. My check was limited because ECHA refused automated access.

Evidence that partly supports the premise. The reason given for the recommendation, that secondary users struggle to control exposure, still has support (see Claim 5).

Verdict: weakened. Policy since 2013 has moved the opposite way, treating beryllium as a strategic material to secure. Exposure control, not elimination, became the accepted answer.

Implications for weight. The recommendation should not be read as a lesson the case establishes. The chapter gave no analysis of substitute performance, of the hazards of alternatives, or of who would bear the cost of substitution. Later policy weighed defence and supply-chain needs and chose control. The transferable point is narrower. Where a hazard cannot be controlled reliably across a dispersed downstream user base, reducing use is a legitimate option to weigh. Whether it is chosen depends on how essential the material is judged to be, and that judgment can harden into strategic designation.

Claim 5: Secondary users and recyclers are unlikely to have the capacity to prevent beryllium disease; downstream, recycling and community exposures continue (pp. 132, 139–140)#

Original claim (pp. 132, 139–140). - Only about 1,500 of 28,000–107,000 exposed US workers work in primary production (p. 132). - A sentinel case at a Quebec metals recycler (1999) led to 31 more cases and a survey finding 2,789 user workplaces (p. 139). - Eight community-acquired CBD cases were recognised in the US in 1999–2002 (p. 139). - It is “unlikely that many secondary users and recyclers have the expertise, resources and knowledge” to prevent disease (p. 140).

Subsequent developments - US regulatory scope followed the life-cycle argument. - OSHA’s feasibility analysis covers twelve application groups. Most are downstream of primary production: foundries, secondary smelting and alloying, precision machining, welding, dental laboratories, aluminium production, coal-fired power and abrasive blasting [S1]. - The final rule extended coverage to construction and shipyards, which the 2015 proposal had not fully included [S1, S10]. - Materials under 0.1% beryllium are exempt only where the employer has objective data showing exposure stays below the action level. OSHA said this was “essentially as proposed by Materion and USW”, acknowledging that trace materials can cause significant exposures [S5]. - The 2020 amendments added labelling requirements for beryllium materials transferred for disposal, recycling or reuse [S7]. - OSHA’s current estimate is about 62,000 exposed workers in about 7,300 establishments [S9]. That sits within the chapter’s earlier range and is not directly comparable to it. - Unexpected and downstream sources since 2013. - Germany (Frye et al. 2021). A CBD case first diagnosed as sarcoidosis led to testing of workmates on a building site. Five of 21 were sensitised. Concrete dust there contained 1,138 μg/kg beryllium, against 147–452 μg/kg at other German sites. The authors describe “a cluster of beryllium-sensitized workers from an industry not related to beryllium” [S36]. A Materion letter disputed the finding. Its title says the finding “requires reexamination”, but I could not access the text [S37]. - Canada (Gravel et al. 2023, IRSST Quebec). In six formal e-waste recycling facilities and one commercial recycler, “dust control was inadequate and personal protective equipment was improperly worn” in most facilities. Beryllium was among the metals quantified in workers’ exposure [S38]. This fits the chapter’s doubts about recyclers’ capacity. - Construction trades at US nuclear sites. 1.2% of 21,854 screened workers were sensitised (1998–2020), and CBD compensation acceptance rose with years of DOE work [S35]. - Clinical review (2026). New cases continue “across both traditional and nontraditional occupational settings” [S42]. - Gaps. I found no post-2013 publication following up the Quebec programme. I also found no new US series of community-acquired CBD cases. Europe PMC searches turned up only analytical-method and pre-2013 Quebec work. The community strand of the claim is therefore neither confirmed nor refuted by new data.

Verdict: strengthened for downstream and recycling exposures, which are now in the US rule’s scope and have new case evidence. Unclear for community exposure.

Implications for weight. Strong. The pattern is well supported: control achievable at the primary producer does not travel down the supply chain, and harm surfaces in sectors that do not know they handle the hazard. It rests on the regulator’s scope decisions and on independent clinical and hygiene studies. The capacity problem among small downstream users and recyclers is documented in at least one systematic field survey [S38].

Claim 6: Interpretation by those with “financial incentives for misinterpretation” must be discounted; product-defence work is “advocacy, rather than science” (p. 140). Guidotti contests this (p. 148)#

Original claims. - Michaels and Monforton (p. 140). “Interpretation of scientific data by those with financial incentives for misinterpretation must be discounted.” Work by product-defence firms “must be seen for what it is: advocacy, rather than science.” - Guidotti (p. 148). Financial and professional incentives attach to “almost everyone”. The real questions are influence, completeness and validity, and excluding corporate research “risks losing an immense body of valuable information”. - The chapter’s example (p. 138). It presents Brush Wellman’s 1999 call for research on particle size and skin exposure as a delaying tactic.

Subsequent developments that support Michaels and Monforton - Industry-linked epidemiology kept contesting lung-cancer risk after 2013. - Boffetta, Fordyce & Mandel (2014) report “lack of an increase in mortality from lung cancer” in workers exposed to insoluble beryllium. Fordyce is at Exponent [S45]. - Their 2016 extension lists Materion Brush Inc. as funder [S46]. - Fordyce et al. (2025) was authored by Exponent staff, a Materion medical director and Boffetta. It reported “no increase in lung cancer or other mortality … at insoluble-only beryllium facilities” [S47]. - Rothman & Mosquin (2013) at RTI Health Solutions argued that the strongest NIOSH association was inflated by “sparse-data bias” [S48]. Their 2017 reanalysis found “modestly but monotonically increasing risk” but confidence bands that “did not clearly separate” low and high exposure [S49]. I could not establish their funding from the metadata I accessed. - NIOSH’s rebuttal. Schubauer-Berigan et al. (2017) analysed two lower-exposure plants (mean 1.3 μg/m³, mainly insoluble beryllium). They found a monotonic exposure-response and concluded the pooled findings “are relevant for current workers exposed to any form of beryllium” [S50]. - OSHA’s treatment. It considered Boffetta’s submissions in detail and “did not find persuasive evidence” that solubility explained the null results. It cited the study’s lack of exposure measurements and the very short employment of most cohort members [S1]. - Industry’s EU positions and corporate language. - In 2017 the EU employers’ group claimed that recent dose-response studies “identify a no-observed adverse effect level (NOAEL) at a higher level” than 0.2 μg/m³. It pushed for 0.6 μg/m³ (inhalable), its own voluntary product-stewardship guideline [S19]. - Materion’s 2025 10-K still says “some scientists claim there is evidence of an association between beryllium exposure and lung cancer”. It refers to “numerous studies affirming the safety of beryllium in these products” [S26]. IARC and the US National Toxicology Program have both classed beryllium as a known human carcinogen since before the chapter (p. 135). - These show that financially interested interpretation stayed systematically slanted on the endpoint still in dispute.

Subsequent developments that support Guidotti - The joint draft standard. In February 2012 Materion and the USW sent OSHA a joint draft standard, which the chapter does not mention. OSHA’s proposal “was based, in part, upon” it, and “many of the provisions in the final rules are identical or substantively similar” [S1]. The shared elements include: - the 0.2 μg/m³ PEL; - the exemption for materials under 0.1% beryllium; - biennial BeLPT surveillance. The regulations.gov record lists it as a letter of 8 February 2012 to David Michaels [S12]. - Collaborative NIOSH–industry studies produced protective findings. - Thomas et al. (2013) documented a prevention programme that worked, with a Materion co-author [S27]. - Virji et al. (2019) studied workers at a primary beryllium manufacturing facility. Peak inhalation metrics and skin-exposure indices were both associated with sensitisation [S31]. - Skin exposure is one of the research questions the chapter treats as a 1999 delaying tactic (p. 138). It later proved real. OSHA applied its PPE requirements broadly “because beryllium sensitization can occur from dermal contact” [S1]. The EU added a dermal sensitisation notation [S17]. This does not show the 1999 request was made in good faith. It does support Guidotti’s point that some of the uncertainty was genuine (p. 148). - OSHA’s practice. The agency weighed industry-sponsored studies on method, not provenance [S1]. That is closer to Guidotti’s auditing model than to discounting.

Where mainstream thinking now stands. Recent work on conflicts of interest in chemicals assessment calls for explicit conflict-of-interest rules and an independent auditor who checks that assessments represent the evidence without bias. It rejects the claim that industry must sit on expert panels because it holds essential data (Schäffer et al. 2023; Michaels is a co-author) [S51]. That position combines both sides of the debate: it treats interested advocacy as a known risk and handles it by auditing rather than by blanket exclusion.

Verdict: partly held up. Interested parties did keep producing slanted interpretations on contested endpoints. But blanket discounting would have discarded the jointly drafted standard and the industry-hosted studies that shaped the rule. Guidotti’s critique is also partly vindicated.

Implications for weight. Use the lesson in a refined form. Expect financially interested interpretations to lean consistently towards the interest, especially on the endpoint still in dispute. Scrutinise their methods and data access harder, rather than setting them aside. The case also shows that the same firm can do both at once: co-author a protective standard where its operations already comply, while contesting carcinogenicity where liability or market exposure remains.

Claim 7: Brush Wellman’s 1977 prediction that carcinogen classification would leave beryllium “no longer a viable industry” was not borne out (pp. 135, 145)#

Original claim. - The chapter quotes Brush Wellman (1977): once classified and regulated as a carcinogen, “it would only be a matter of time until its usage would shrink to a point where it would no longer be a viable industry” (p. 135). - Beryllium was nonetheless classified: IARC Group 1 in 1994 and 2009, and the US National Toxicology Program’s “known human carcinogen” in 2002 (p. 135). - Guidotti describes the firm as “still” highly profitable with a “near monopoly” (p. 145).

Subsequent developments - Financial trajectory. Materion’s net sales were $1,786.6 million in 2025 and $1,684.7 million in 2024. Net income was $95.7 million in 2023, $5.9 million in 2024 and $74.8 million in 2025 [S26]. In 2012 net income was $24.7 million [S24]. The company employs about 2,880 people. It operates what it calls “the world’s largest bertrandite ore mine and refinery”. Bertrandite is the beryllium ore. - Business mix. The firm has diversified, but beryllium remains core. The Performance Materials segment, its beryllium business, had 2025 net sales of $675.9 million and EBITDA of $127.2 million [S26]. - Costs of the lower limit. OSHA judged the 2017 rule economically feasible in all sectors, at an estimated $74 million a year across all employers [S1]. The firm said the rule reflected its existing practices [S25]. - Litigation tail. Materion reported no pending beryllium lawsuits at 31 December 2011, one at 31 December 2012 [S24], one (with four plaintiffs) at 31 December 2017 [S25], and none at 31 December 2025 [S26]. - State support (see Claim 4). It includes Defense Production Act co-funding, critical-mineral status in the US and EU, and a federal production tax credit [S20, S22–S24, S26].

Caveat. The industry’s survival owed much to defence demand and state support that the 1977 forecast did not consider. The forecast was wrong about outcomes. That does not show carcinogen classification was commercially harmless in general. Materion itself still lists health concerns and substitution as risks [S26].

Verdict: held up. The prediction was not borne out, and nothing since 2013 changes that.

Implications for weight. The case adds to the pattern of industry forecasts of ruin from hazard classification or tighter limits that did not materialise. The evidence is suggestive, not general. This is one firm, protected by strategic demand and near-monopoly supply, so its survival says little about more competitive markets. Use it to discount such forecasts moderately, not to dismiss them.

Claim 8: Companies will change course only if given exit routes (forgiven legacy liability, due-diligence defences, a “threshold for sufficient knowledge”); “there must be room for them to turn around” (pp. 149–150)#

Original claim (pp. 149–150). - Guidotti argues that “organisations cannot be expected to change their positions unless they are given a ‘way out’”, which may involve “forgiving past liability and reducing punitive damages”. - He notes, “This is largely what happened with the DOE contract worker’s compensation programme.” - He also points to the threat of shareholder or board action as a counterweight.

Subsequent developments - The compensation programme’s exit route (legislated before 2013). Under the Energy Employees Occupational Illness Compensation Program Act (EEOICPA), 42 U.S.C. §7385d, eligibility for federal compensation depends on dropping or not pursuing “a tort case … against a beryllium vendor or atomic weapons employer” [S16]. That is a statutory channel that diverts claims from vendors such as Brush Wellman. So Guidotti’s parenthesis is accurate. - Order of events. - Materion reported no pending beryllium lawsuits by 31 December 2011 [S24]. - About six weeks later, on 8 February 2012, Materion and the USW sent OSHA their joint draft standard with a tenfold-lower PEL [S12, S1]. - The litigation tail stayed minimal afterwards [S25, S26]. The firm’s public turn thus came once its legacy liability had largely run off, which fits Guidotti’s mechanism. - What the evidence does not show. - No formal forgiveness, safe harbour or due-diligence defence accompanied the 2017 rule [S1]. - The joint proposal also matched the firm’s commercial position. Materion said the resulting standard “fundamentally represent[s] our current health and safety operating practices” [S25]. It said it “was a participant in the development of the new standards”. A uniform standard also puts competitors and downstream users on the same footing. - I found no primary source stating the firm’s motives. The 2012 letter itself was not retrievable (regulations.gov 403) [S12]. - Limits of the turn. The company went on contesting other points: carcinogenicity [S26], the European limit [S19] and a sensitisation cluster [S37]. Its turn was selective, not a general shift in posture.

Verdict: partly held up. The sequence fits the exit-route thesis: liability extinguished (partly by statute), then cooperation. But the evidence is circumstantial, and a simpler explanation fits equally well. The firm backed a limit it already met.

Implications for weight. Treat this as a suggestive mechanism, not an established one. The case supports the narrower proposition that settling legacy liability, for example through a compensation scheme that channels claims away from tort, can remove a barrier to cooperation on forward-looking standards. It does not show that exit routes are necessary. It also shows that interest-alignment is a competing explanation that must be ruled out.

Claim 9: Standards become “sticky”, so a policy of continuous improvement is preferable, “although perhaps unattainable in the current political context” (p. 150)#

Original claim (p. 150). - “Once a standard is set, it becomes ‘sticky’. It develops a constituency and an infrastructure to support it.” - “A policy of sequential standard-setting based on scientific evidence is inherently flawed”. It is better “to accept a realistic policy of continuous improvement, anticipating that standards evolve”.

Subsequent developments: stickiness - OSHA. The 2 μg/m³ limit, set tentatively in 1949 and adopted by OSHA in 1971–72, was replaced in January 2017 and enforced from May 2018 [S1, S7]. - DOE. - DOE’s 1999 Chronic Beryllium Disease Prevention Program was the chapter’s model of early action (p. 138). - In 2016 DOE proposed lowering its action level to 0.05 μg/m³ [S13]. - In 2023 it offered instead an action level of 0.1 μg/m³ and its own PEL of 0.2 μg/m³, both matching OSHA’s numbers, and a 2.0 μg/m³ STEL [S14]. That pulled back from its earlier proposal to OSHA’s new benchmark, which is itself an anchoring effect. - As of the Fall 2025 Unified Agenda the rule is a “Long-Term Action” with final action projected for July 2027. The Spring 2025 agenda had projected September 2026 [S15]. - No final rule appears in the Federal Register through September 2026 [S11]. The 1999 rule has now stood 27 years. - EU. - SCOEL recommended 0.02 μg/m³ (inhalable) [S19]. - The tripartite ACSH agreed 0.2 μg/m³, with 0.6 μg/m³ for a five-year transition, because employers found 0.2 “very challenging” [S19]. - The Commission proposed five years [S18]. The final Directive (EU) 2019/983 set seven: “0,0006 mg/m³ until 11 July 2026” [S17]. - Its recitals ask the Commission to review limits “on a regular basis” [S17]. I found no proposal to revisit beryllium. - The ACGIH’s 0.05 μg/m³ (2009, cited by OSHA [S1]) remains below both binding limits.

Subsequent developments: staged and continuous approaches - Staged approaches were used: - OSHA’s phased compliance dates for engineering controls (2020) [S7]; - an action level at half the PEL that triggers surveillance [S1]; - the EU’s transitional value [S17]. - These are one-off glide paths to a new fixed number. They are not a standing commitment to keep tightening. None of the regimes I examined builds in a scheduled ratchet or automatic review for beryllium. DOE’s 2016 plan to adopt “any future” OSHA PEL automatically, which would have been a modest link to future tightening, was replaced in 2023 by its own fixed PEL [S13, S14].

Verdict: strengthened on stickiness. On continuous improvement, the prescription has not been adopted, which fits Guidotti’s own “perhaps unattainable”.

Implications for weight. Strong for the descriptive mechanism. The events after 2013 add two refinements: - A new number becomes the next anchor. DOE, which was ahead in 1999, converged down to OSHA’s 2017 values rather than beyond them. - Transitional periods themselves get negotiated longer, from five years to seven in the EU.

The continuous-improvement alternative remains untested for this substance.

Claim 10: CBD risk is probably driven by short peak exposures that standard full-shift measurement misses (p. 134)#

Original claim (p. 134). “With hindsight, we can speculate” that CBD is immune-mediated, with “considerable inter-individual variability in susceptibility, and a dose-response relationship which is probably driven by ‘peak’ exposures — possibly of very short duration — which standard methods of exposure assessment do not detect.” The authors flag this as speculation.

Subsequent developments - Supportive but not decisive. - Virji et al. (2019, NIOSH) studied 264 short-term workers (1994–1999) at a primary beryllium manufacturing facility. Metrics of peak inhalation exposure and indices of skin exposure were both significantly associated with sensitisation. “However … we could not tease apart the independent effects of skin exposure from inhalation exposure.” They recommend controlling airborne exposure “with attention to peaks” and eliminating skin contact [S31]. - Harber et al. (2014) analysed 532 participants in the Beryllium BioBank, a research cohort of exposed workers. Progression from sensitisation to CBD was associated with “peak-level weighted exposure hours” [S29]. - A NIOSH review of peak-exposure metrics found “consensus is lacking on its definition” (Virji & Kurth 2020) [S32]. - Weak exposure-response for sensitisation [S33, S34]. This is consistent with a peak or skin route that full-shift averages miss. It is also consistent with poor reconstruction of historical exposure. - Individual susceptibility, the chapter’s other element, is now firmly established through the HLA-DPB1 E69 genotype [S28, S30, S34]. - Do short-term limits address it? - OSHA set a 15-minute STEL of 2.0 μg/m³, ten times the PEL. It said “the available epidemiological literature … does not address” whether a 2.0 or 1.0 μg/m³ STEL would differ, because “detailed documentation of workers’ short-term exposures is typically not available to researchers”. So it relied on reducing excursions and on feasibility [S1]. - OSHA also noted that “consistent with the comments from Materion, the identification and control of short-term exposures is critical” [S1]. Unions and National Jewish Health had pressed for 0.5–1.0 μg/m³ [S1]. - SCOEL recommended a 15-minute STEL of 0.2 μg/m³ (inhalable) [S19]. The EU directive set no STEL for beryllium [S17]. - DOE’s 2023 proposal would add a 2.0 μg/m³ STEL [S14].

Verdict: partly held up. Peak exposure remains a plausible and partly supported contributor. It has not been shown to be the driver. Skin contact, which the chapter’s hypothesis does not mention, has emerged as a co-equal route for sensitisation. Current short-term limits are set on feasibility, not on a peak-based dose-response.

Implications for weight. Use the general insight with confidence: the metric a standard regulates, here the full-shift airborne average, can be misaligned with the biological mechanism, and a standard that is “met” can still leave the relevant dose uncontrolled. The later evidence reinforces this insight while changing the specific mechanism the chapter proposed. It widens it: the regulated metric missed both short peaks and a whole exposure route, the skin.

Summary of verdicts#

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

Each is tied to the section’s pages and to the later evidence above.

1. Provisional numbers harden into anchors, and replacing one creates the next anchor. A limit adopted “tentatively” (p. 133) stood for about seven decades. When it was replaced, other institutions converged on the new number rather than on independent scientific advice. Transitions were negotiated longer (pp. 133, 150; [S1, S13–S15, S17–S19]).
2. The regulated metric can be misaligned with the harm mechanism. Compliance with a full-shift airborne average did not control short peaks or skin contact. What worked was multi-route control (pp. 134, 139; [S1, S27, S31]).
3. Residual risk at the chosen limit can be openly acknowledged, and the stopping point set by feasibility rather than safety. That makes the trade-off visible and calls for ancillary measures such as action levels, surveillance and hygiene (p. 139; [S1]).
4. Control achievable by the lead producer does not travel down a dispersed supply chain. Harm surfaces among secondary users, recyclers and sectors unaware of the hazard (pp. 132, 139–140; [S1, S35, S36, S38]).
5. Interested-party science tends to stay slanted on the endpoint still in dispute, even while the same party cooperates where it already complies. This argues for auditing method and data, not for blanket inclusion or exclusion (pp. 140, 148; [S1, S12, S45–S51]).
6. Settling legacy liability, for example by channelling claims into a compensation scheme, may remove a barrier to forward-looking cooperation. Interest-alignment is an equally plausible explanation and must be tested (pp. 149–150; [S16, S24–S26]).
7. Forecasts of industry ruin from hazard classification can prove wrong, especially where the material is strategically protected. Survival may depend on state support rather than on the harmlessness of regulation (pp. 135, 145; [S21–S26]).
8. Long-latency, often-misdiagnosed harms defeat quick evaluation of interventions. Projected benefits can only be checked if surveillance data are built in and pooled (p. 140; [S1, S39, S44]).
9. Strategic designation can override a substitution agenda. Once a hazardous material is classed as critical to security or supply, policy turns from reducing use to securing it, and control becomes the default (pp. 139–140; [S20, S22–S24, S26]).

Sources#

All retrieved 25 September 2026 unless noted.

US regulation (Federal Register, OSHA, DOE, statute) - [S1] OSHA. Occupational Exposure to Beryllium; Final Rule. 82 FR 2470–2757, 9 January 2017 (signed by David Michaels, 14 December 2016). https://www.federalregister.gov/documents/2017/01/09/2016-30409/occupational-exposure-to-beryllium. Full text read at https://www.govinfo.gov/content/pkg/FR-2017-01-09/html/2016-30409.htm (pages cited: 2470–2471, 2478, 2528–2535, 2542–2543, 2553–2554, 2618–2619, 2698). - [S2] OSHA. Occupational Exposure to Beryllium: Delay of Effective Date. 82 FR 8901, 1 February 2017. https://www.federalregister.gov/documents/2017/02/01/2017-02149/occupational-exposure-to-beryllium-delay-of-effective-date - [S3] OSHA. Occupational Exposure to Beryllium; Further Delay of Effective Date. 82 FR 14439, 21 March 2017. https://www.federalregister.gov/documents/2017/03/21/2017-05569/occupational-exposure-to-beryllium-further-delay-of-effective-date - [S4] OSHA. Occupational Exposure to Beryllium and Beryllium Compounds in Construction and Shipyard Sectors (proposed rule to revoke ancillary provisions). 82 FR 29182, 27 June 2017. https://www.federalregister.gov/documents/2017/06/27/2017-12871/occupational-exposure-to-beryllium-and-beryllium-compounds-in-construction-and-shipyard-sectors - [S5] OSHA. Revising the Beryllium Standard for General Industry (direct final rule). 83 FR 19936, 7 May 2018 (effective 6 July 2018; confirmed 83 FR 31045, 3 July 2018). https://www.federalregister.gov/documents/2018/05/07/2018-09306/revising-the-beryllium-standard-for-general-industry - [S6] OSHA. Occupational Exposure to Beryllium and Beryllium Compounds in Construction and Shipyard Sectors (final rule declining to revoke ancillary provisions; compliance dates). 84 FR 51377, 30 September 2019. https://www.federalregister.gov/documents/2019/09/30/2019-21037/occupational-exposure-to-beryllium-and-beryllium-compounds-in-construction-and-shipyard-sectors - [S7] OSHA. Revising the Beryllium Standard for General Industry (final rule). 85 FR 42582, 14 July 2020 (effective 14 September 2020). https://www.federalregister.gov/documents/2020/07/14/2020-10678/revising-the-beryllium-standard-for-general-industry. Full text: https://www.govinfo.gov/content/pkg/FR-2020-07-14/html/2020-10678.htm - [S8] OSHA. Occupational Exposure to Beryllium and Beryllium Compounds in Construction and Shipyard Sectors (final rule). 85 FR 53910, 31 August 2020 (effective 30 September 2020). https://www.federalregister.gov/documents/2020/08/31/2020-18017/occupational-exposure-to-beryllium-and-beryllium-compounds-in-construction-and-shipyard-sectors. Full text: https://www.govinfo.gov/content/pkg/FR-2020-08-31/html/2020-18017.htm - [S9] OSHA. Beryllium (safety and health topic page: about 62,000 exposed workers in about 7,300 establishments). https://www.osha.gov/beryllium - [S10] OSHA. Occupational Exposure to Beryllium and Beryllium Compounds (proposed rule). 80 FR 47566, 7 August 2015. https://www.federalregister.gov/documents/2015/08/07/2015-17596/occupational-exposure-to-beryllium-and-beryllium-compounds - [S11] Federal Register API searches for “beryllium” (all agencies, 1 June 2024 to 25 September 2026; and OSHA, all dates), run 25 September 2026. https://www.federalregister.gov/api/v1/documents.json?conditions[term]=beryllium&conditions[publication_date][gte]=2024-06-01 - [S12] Materion Brush Inc. and United Steelworkers. Letter of 8 February 2012 to David Michaels (OSHA) with a joint draft recommended beryllium standard. Docket OSHA-H005C-2006-0870, document 0754 (metadata via regulations.gov API; PDF refused, HTTP 403). https://www.regulations.gov/document/OSHA-H005C-2006-0870-0754 - [S13] US Department of Energy. Chronic Beryllium Disease Prevention Program (notice of proposed rulemaking). 81 FR 36704, 7 June 2016 (DOE registry data through 2013, Table 1). https://www.federalregister.gov/documents/2016/06/07/2016-12547/chronic-beryllium-disease-prevention-program - [S14] US Department of Energy. Chronic Beryllium Disease Prevention Program (supplemental notice of proposed rulemaking). 88 FR 57365, 23 August 2023. https://www.federalregister.gov/documents/2023/08/23/2023-18082/chronic-beryllium-disease-prevention-program - [S15] Office of Information and Regulatory Affairs. Unified Agenda entry, DOE RIN 1992-AA39, Chronic Beryllium Disease Prevention Program. Spring 2025 (final action projected 09/2026): https://www.reginfo.gov/public/do/eAgendaViewRule?pubId=202504&RIN=1992-AA39. Fall 2025 (“Long-Term Actions”, final action projected 07/2027): https://www.reginfo.gov/public/do/eAgendaViewRule?pubId=202510&RIN=1992-AA39 - [S16] 42 U.S.C. §7385d, Election of remedy for beryllium employees and atomic weapons employees (Energy Employees Occupational Illness Compensation Program Act, as amended 2001). https://uscode.house.gov/view.xhtml?req=granuleid:USC-prelim-title42-section7385d&num=0&edition=prelim

EU - [S17] Directive (EU) 2019/983 of 5 June 2019 amending Directive 2004/37/EC (carcinogens and mutagens at work), recitals 19–20 and 29, and Annex III entry for beryllium (0.0002 mg/m³ inhalable; 0.0006 mg/m³ until 11 July 2026; dermal and respiratory sensitisation). https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32019L0983 (I could not confirm the current consolidated text; EUR-Lex refused later requests.) - [S18] European Commission. Proposal COM(2018) 171 final, 5 April 2018 (2018/0081(COD)), recital 14 and explanatory memorandum. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52018PC0171 - [S19] Advisory Committee on Safety and Health at Work (ACSH). Opinion on an EU Occupational Exposure Limit value for Beryllium and Inorganic Beryllium Compounds under Directive 2004/37/EC. Doc. 662/17, adopted 31 May 2017 (includes SCOEL Recommendation 175 of 8 February 2017, and employer and worker group statements). https://circabc.europa.eu/sd/a/2d61770f-7b5d-45bc-b2b4-4c8460e78c93/Doc.662-17-EN_WPC_Opinion%20on%20Be_Adopted%2031.05.2017.pdf - [S20] European Commission, DG GROW. Critical raw materials (fifth list, 2023; beryllium included). https://single-market-economy.ec.europa.eu/sectors/raw-materials/areas-specific-interest/critical-raw-materials_en

Supply, industry and company filings - [S21] US Geological Survey. Mineral Commodity Summaries 2013: Beryllium (January 2013). https://d9-wret.s3.us-west-2.amazonaws.com/assets/palladium/production/mineral-pubs/beryllium/mcs-2013-beryl.pdf - [S22] US Geological Survey. Mineral Commodity Summaries 2026: Beryllium (February 2026), pp. 52–53. https://pubs.usgs.gov/periodicals/mcs2026/mcs2026-beryllium.pdf (full volume https://pubs.usgs.gov/periodicals/mcs2026/mcs2026.pdf) - [S23] US Department of the Interior / USGS. Final 2025 List of Critical Minerals. 90 FR 50494, 7 November 2025. https://www.federalregister.gov/documents/2025/11/07/2025-19813/final-2025-list-of-critical-minerals - [S24] Materion Corporation. Form 10-K for fiscal year 2012 (filed 8 March 2013): Item 3, Beryllium Claims; Defense Production Act Title III project; net income. https://www.sec.gov/Archives/edgar/data/1104657/000110465713000047/mtrn-20121231x10k.htm - [S25] Materion Corporation. Form 10-K for fiscal year 2017 (filed 15 February 2018): OSHA standard “fundamentally represent[s] our current health and safety operating practices”; Beryllium Claims. https://www.sec.gov/Archives/edgar/data/1104657/000110465718000007/mtrn_2017123110k.htm - [S26] Materion Corporation. Form 10-K for fiscal year 2025 (filed 12 February 2026): risk factors on CBD, lung cancer and substitutes; segment results; DoD reimbursement; Advanced Manufacturing Production Credit; Item 3 (no pending beryllium cases). https://www.sec.gov/Archives/edgar/data/1104657/000110465726000011/mtrn-20251231.htm

Peer-reviewed science (abstracts and metadata via Europe PMC and Crossref unless noted) - [S27] Thomas, C. A., Deubner, D. C., Stanton, M. L., Kreiss, K., Schuler, C. R. Long-term efficacy of a program to prevent beryllium disease. Am J Ind Med 2013;56:733–741. https://doi.org/10.1002/ajim.22175 - [S28] Balmes, J. R., et al. (ATS Ad Hoc Committee on Beryllium Sensitivity). An official American Thoracic Society statement: diagnosis and management of beryllium sensitivity and chronic beryllium disease. Am J Respir Crit Care Med 2014;190:e34–e59. https://doi.org/10.1164/rccm.201409-1722st - [S29] Harber, P., Su, J., Alongi, G. Exposure factors associated with chronic beryllium disease development in Beryllium BioBank participants. J Occup Environ Med 2014;56:852–856. https://doi.org/10.1097/jom.0000000000000190 - [S30] Kreiss, K., Fechter-Leggett, E. D., McCanlies, E. C., Schuler, C. R., Weston, A. Research to practice implications of high-risk genotypes for beryllium sensitization and disease. J Occup Environ Med 2016;58:855–860. https://doi.org/10.1097/jom.0000000000000805 - [S31] Virji, M. A., et al. Associations of metrics of peak inhalation exposure and skin exposure indices with beryllium sensitization at a beryllium manufacturing facility. Ann Work Expo Health 2019;63:856–869. https://doi.org/10.1093/annweh/wxz064 - [S32] Virji, M. A., Kurth, L. Peak inhalation exposure metrics used in occupational epidemiologic and exposure studies. Front Public Health 2020;8:611693. https://doi.org/10.3389/fpubh.2020.611693 - [S33] Widyaningsih, V., et al. Lack of an exposure response and interaction with HLA-DPβ1 and DRβ1 polymorphisms in the development of beryllium toxicity in a high beryllium exposure cohort. J Occup Environ Med 2020;62:64–73. https://doi.org/10.1097/jom.0000000000001766 - [S34] Crooks, J., et al. HLA-DPB1 E69 genotype and exposure in beryllium sensitisation and disease. Occup Environ Med 2022;79:120–126. https://doi.org/10.1136/oemed-2021-107736 - [S35] Cloeren, M., et al. Beryllium disease among construction trade workers at Department of Energy nuclear sites: a follow-up. Am J Ind Med 2022;65:708–720. https://doi.org/10.1002/ajim.23411 - [S36] Frye, B. C., et al. A cluster of beryllium sensitization traced to the presence of beryllium in concrete dust. Chest 2021;159:1084–1093. https://doi.org/10.1016/j.chest.2020.09.073 - [S37] Kolanz, M., Dumas, C. (Materion). Chronic beryllium disease finding requires reexamination (letter). Chest 2021;159:2508–2509. https://doi.org/10.1016/j.chest.2020.12.061 (title and affiliation only; text not accessible) - [S38] Gravel, S., et al. Occupational health and safety, metal exposures and multi-exposures health risk in Canadian electronic waste recycling facilities. Waste Manag 2023;165:140–149. https://doi.org/10.1016/j.wasman.2023.04.026 - [S39] Mayer, A. S., et al. Sarcoidosis in beryllium exposed workers: a case-case study. Am J Ind Med 2025;68:68–75. https://doi.org/10.1002/ajim.23676 - [S40] Maier, L. A., Mroz, M. M., Lin, N., Mayer, A., Barker, E. Sarcoidosis in the military may be chronic beryllium disease (letter). Chest 2024;165:e25. https://doi.org/10.1016/j.chest.2023.07.4221 (title only) - [S41] MacMurdo, M. G., et al. Chronic beryllium disease: update on a moving target. Chest 2020;158:2458–2466. https://doi.org/10.1016/j.chest.2020.07.074 - [S42] Morgan, A., MacMurdo, M. G. New aspects of chronic beryllium disease. Curr Opin Allergy Clin Immunol 2026;26:100–106. https://doi.org/10.1097/aci.0000000000001140 - [S43] TenHarmsel, H., Wang, L., Dumas, C., Rosenman, K. D. Mortality among individuals diagnosed with chronic beryllium disease and beryllium sensitization. J Occup Environ Med 2024;66:247–251. https://doi.org/10.1097/jom.0000000000003033. With Boffetta, P., letter, J Occup Environ Med 2024;66:e323, https://doi.org/10.1097/jom.0000000000003097, and authors’ reply, https://doi.org/10.1097/jom.0000000000003098 - [S44] Bell, J. L., Mazurek, J. M. Trends in pneumoconiosis deaths — United States, 1999–2018. MMWR 2020;69(23):693–698. https://www.cdc.gov/mmwr/volumes/69/wr/mm6923a1.htm - [S45] Boffetta, P., Fordyce, T., Mandel, J. S. A mortality study of workers exposed to insoluble forms of beryllium. Eur J Cancer Prev 2014;23:587–593. https://doi.org/10.1097/cej.0000000000000013 - [S46] Boffetta, P., Fordyce, T. A., Mandel, J. S. A mortality study of beryllium workers. Cancer Med 2016;5:3596–3605 (funder listed: Materion Brush Inc). https://doi.org/10.1002/cam4.918 - [S47] Fordyce, T. A., Leonhard, M. J., Jung, A. M., Lin, J., Watson, H., Heldt, G. H., Boffetta, P. An updated mortality study of beryllium workers, 1925–2020. J Occup Environ Med 2025;67:e679–e689 (affiliations: Exponent, Inc.; Materion Brush, Inc.; Stony Brook). https://doi.org/10.1097/jom.0000000000003463 - [S48] Rothman, K. J., Mosquin, P. L. Sparse-data bias accompanying overly fine stratification in an analysis of beryllium exposure and lung cancer risk. Ann Epidemiol 2013;23:43–48. https://doi.org/10.1016/j.annepidem.2012.11.005 - [S49] Mosquin, P. L., Rothman, K. J. Reanalysis of reported associations of beryllium and lung cancer in a large occupational cohort. J Occup Environ Med 2017;59:274–281. https://doi.org/10.1097/jom.0000000000000947 - [S50] Schubauer-Berigan, M. K., Couch, J. R., Deddens, J. A. Is beryllium-induced lung cancer caused only by soluble forms and high exposure levels? Occup Environ Med 2017;74:601–603. https://doi.org/10.1136/oemed-2016-104064 - [S51] Schäffer, A., et al. (including D. Michaels). Conflicts of interest in the assessment of chemicals, waste, and pollution. Environ Sci Technol 2023;57:19066–19077. https://doi.org/10.1021/acs.est.3c04213 - [S52] Devoy, J., Remy, A. M., La Rocca, B., Wild, P., Rousset, D. Occupational exposure to beryllium in French industries. J Occup Environ Hyg 2019;16:229–241. https://doi.org/10.1080/15459624.2018.1559926