On the basis of a hearing before an administrative magistrate (AM), the plaintiff’s claim for accidental disability retirement benefits was denied by the Contributory Retirement Appeal Board (CRAB). In this appeal from a Superior Court judgment upholding the CRAB decision, the plaintiff asks us to rule, as matter of law, that she established that the air quality in her work place proximately caused her to be disabled. We decline to do so, and uphold the ruling of CRAB.
*128Plaintiff’s claim. Beginning in September of 1993, the plaintiff, a music teacher in the Natick public schools, complained of headaches, head congestion, and, eventually, respiratory difficulties. She began treatment with her physician for sinusitis, without relief. From November 12, 1993, to January 3, 1994, she took sick leave (except for the holidays, which did not count as sick leave). At that time she was diagnosed with viral bronchitis. The plaintiff thereafter began a long series of consultations with various clinicians, who came to support her claim of work-related illness. She stopped going to work after returning for one half day on March 1, 1994, and began to receive workers’ compensation benefits under G. L. c. 152. Eventually, on December 7, 1995, she filed a claim for accidental disability retirement under G. L. c. 32, § 7(1), with the Teachers’ Retirement Board (TRB). After a hearing before the AM, the plaintiff’s claim was denied. Her appeal to CRAB followed. CRAB adopted the AM’s decision as its own, and CRAB’s decision was affirmed on further appeal to the Superior Court.
The main difficulty with the plaintiff’s claim was to demonstrate the requisite causal connection between her medical condition and the environment in which she worked. The plaintiff was required to prove her disablement was “by reason of ... a hazard undergone as a result of, and while in the performance of, [her] duties” of employment. Blanchette v. Contributory Retirement Appeal Bd., 20 Mass. App. Ct. 479, 482 (1985), quoting from G. L. c. 32, § 7(1), as appearing in St. 1982, c. 630, § 18.2 That is, the plaintiff’s proof had to satisfy CRAB that her disability was the “natural and proximate result” of the hazard undergone. Campbell v. Contributory Retirement Appeal Bd., 17 Mass. App. Ct. 1018, 1018-1019 (1984), quoting from G. L. c. 32, § 6(3)(a), as amended by St. 1946, c. 603, § 2. See Blanchette v. Contributory Retirement Appeal Bd., supra at 483, 485. To succeed, therefore, she had to establish that her condition was caused by environmental factors in the work place and not due to some other common etiology. The plaintiff also had to rule out environmental factors *129other than her school environment or to find some other way of directly linking her school environment to her illness. In the context of this claim (which is for disability based upon a gradual deterioration at work, and not upon some sudden traumatic event), she had to show as part of her case that she was exposed at work to a hazard “not common and necessary to all or a great many occupations.” Blanchette v. Contributory Retirement Appeal Bd., supra at 485, quoting from Kelly’s Case, 394 Mass. 684, 688 (1985). The plaintiff had the burden of proof on the causation issue. Campbell v. Contributory Retirement Appeal Bd., supra at 1019. Lisbon v. Contributory Retirement Appeal Bd., 41 Mass. App. Ct. 246, 255 (1996). In order to understand how the plaintiff approached the task of meeting this burden, we summarize the evidence before the AM that bears on the issue of causation.
The plaintiff testified at the hearing before the AM as to her job activities and the environment in which she worked. As an elementary school music teacher, she was required to speak and demonstrate songs throughout the school day. She would have time at the beginning, the middle, and the end of the school day to prepare. Because of increased enrollments she lost her own classroom at one of the schools at which she taught, the BennettHemenway School, known as “the Ben-Hem.” There she was required to conduct her office activities in a basement space that she said was musty and had poor ventilation. This change of location took place in the fall of 1993. She had no complaints about her working environment in the other school at which she taught, the Lilja School.
Her main evidence on causation was that her respiratory condition deteriorated markedly after she began using the basement office at the Ben-Hem, but that her condition improved when she was not exposed to the environment in that basement space. This would be consistent with a causal connection between her work environment and her physical condition. Other evidence in the record, however, fairly detracts from this conclusion. First, although she manifested symptoms of some respiratory illness when she was seen by a doctor in December, 1995, and January, 1996, the doctor was unable to verify loss of bronchial or pulmonary function at that time by clinical testing. *130Her spirometry tests were within normal limits at that time. She did not respond positively to medications that would relieve such claimed conditions. The medical records in evidence noted that her complaints at this time were subjective and, in some sense, paradoxical.
To pursue the possibility of an allergic reaction mechanism the plaintiff underwent several tests. As a result she was shown to be allergic to house dust, mites, dog and cat dander, and certain mold spores. But these types of allergens are not uncommon in the environment.
The plaintiff also had a personal medical history of respiratory problems. She had been treated for many years for sinusitis, which she developed after surgery for a deviated septum in 1989. From 1989 to 1993, she experienced recurring sinus headaches, pressure, and discomfort, usually in the fall and spring. In the spring of 1993, she was treated for sinusitis. Her family history disclosed that her mother and brother had a history of asthma.
She importuned one of her physicians to support her claim of a work-related disability, and he was sympathetic. However, he advised her that her claim would be difficult to prove unless she could establish that particular allergens were present at her work place and not present in her home or elsewhere in the environment to which she was regularly exposed.
At the plaintiff’s request the school department commissioned an air quality study by Universal Engineering Corporatian (Universal Engineering), an environmental testing firm that examined both her work space and her home. The report of the testing was inconclusive. It failed to confirm that mold spores to which she was allergic were present in significant concentratians at her work place but not at her home. Thus her claim was not advanced by this report.
The plaintiff’s case relating to causation thus rested on the following factual premises and logic: (1) she actually had respiratory difficulties closely related in time to her exposure to her work space at the Ben-Hem; (2) her self-reporting of these symptoms was reliable and significant, even though not clinically confirmed by physical examinations or objective testing close in time to the claimed effect; (3) her prior history of *131respiratory illness did not explain her complaints; (4) her family history of asthma and related respiratory issues was not significant; and (5) any respiratory difficulties that occurred or persisted when she was not in contact with the Ben-Hem were not significant or were explained by other causes.
Expert opinion evidence. The AM admitted into evidence the medical records of the plaintiff and other documents in which various physicians commented on her condition and treatment and offered opinions as to the nature of her illness, her impairments, her prognosis, and the causes of her condition.3
Several doctors offered their written opinions that the plaintiff’s disability was caused by her work environment. These supportive opinions credited the plaintiff’s self-reporting of her symptoms and found significant that, in their view, any evidence of other identified sources of allergens or causes outside of her school environment was lacking.
It should be noted that not all of these opinions were unqualified or couched in terms of a reasonable medical certainty.4 For example, Dr. Andrew Davidson stated, “I suspect that [the plaintiff and her husband] are correct in their presumed diagnosis of occupationally induced asthma from molds in the school building ...” (emphasis supplied). “It is still my belief that this patient is suffering from occupational asthma as a cause of her symptoms” (emphasis supplied).
Doctor Michael Lawrence stated, “Now that we have confirmed the diagnosis of asthma with a positive Methacholine Challenge, I do feel that there is a definite relationship between exacerbation of [the plaintiff’s] symptoms and exposure to whatever seems to be causing her difficulty [at the Ben-Hem]” (emphasis supplied).
Doctor David Christiani opined, “The time course of [the plaintiff’s] illness and recovery pattern is very typical in my opinion of occupational related asthma and I have no reason to *132doubt that despite the fact that a single agent may be difficult to identify, this represents an occupational related asthma condition” (emphasis supplied).
Doctor Ronald Dorris observed, “My diagnosis is Asthma and Rhinitis, Occupational, due to or aggravated by employment in the [Ben-Hem]. The causal relationship is based on my examination and the history provided to me. My reason for this causal relationship opinion is based on extensive experience with such patients and the accurate ring of [the plaintiff’s] history and the sequence of events, with appropriate documentation of [her] symptoms through spirometric testing.”5 (Emphasis supplied.) Doctor Dorris added, “[The plaintiff’s] excellent [i.e., thorough] history supports of [sic] the relation of her symptoms to her work site. As symptoms and respiratory reactivity worsen they usually spread so that additional locations and exposures cause symptoms. Therefore the initial history is very important. Hers clearly incriminates the lowest floor of the [Ben-Hem]. . . . If [the plaintiff’s] history is accepted as reliable, particularly if it is reinforced by similar complaints from other employees,[6] then the relationship between [the plaintiff’s] symptoms and her exposures [at the Ben-Hem] should be accepted.” (Emphases supplied.)
The record also contained the certificates and narrative reports6 7 of the three physicians who comprised the regional medical panel (panel) that considered the plaintiff’s claim for accidental disability retirement, as required by G. L. c. 32, § 6(3)(a). See Fairbairn v. Contributory Retirement Appeal Bd., 54 Mass. App. Ct. 353, 354-357 (2002), for a description of the regional medical panel, the certificate form, and the function it serves. Here, the panel members each answered in the af*133firmative the three questions on the certificate form regarding, respectively, (1) extent of incapacity, (2) permanency of incapacity, and (3) possibility of causation.8 Subsequent to the submission of the panel members’ reports, the TRB submitted a request for clarification to each of the three physicians pursuant to 840 Code Mass. Regs. § 10.11(2) (1998). See, e.g., Houde v. Contributory Retirement Appeal Bd., 57 Mass. App. Ct. 842, 845, 850 n.10 (2003).9 Among the questions posed to each doctor by the TRB in its request for clarification was the following:
“Please elaborate on the significance of the results from engineering air quality study done on both the [plaintiff’s] home and work environments. Since viable fungus spores were found in both the home and work environments, can you determine which source contributed primarily to the causation and continuation of the [plaintiff’s] condition?”10
Each of the three doctors on the panel made statements in their narratives that were helpful to the plaintiff’s case. On the other hand those narratives, as supplemented by their clarifications, also contained information that could have tended to weaken the plaintiff’s claim. For example, Dr. William Fishbaugh, Jr., noted that despite the plaintiff’s move to New *134Hampshire her symptoms have not changed or gotten worse, and that she has problems because she had other allergies. He also noted that her mother had asthma. In his clarification, Dr. Fishbaugh stated, “[WJhether [it was] the home environment or the work environment . . . which was the major contributor, it’s spores, and it doesn’t make any difference if it is at home or at work.’’ Doctor Thomas Morris, III, originally noted that the plaintiff had normal spirometry in lanuary, 1994, and that she had a paradoxically negative response to bronchodilators. In his clarification he noted that he could not determine which source, as between home and work, contributed primarily to the causation and continuation of her problems. He also stated that if she were not in an environment that exposed her to mold spores there was no reason she could not perform the essential duties of her job. (Only the Ben-Hem was shown to have had mold spores, and there was no evidence that other schools in the district had similar problems). In his original narrative Dr. Ronald Sen noted that the plaintiffs nature of presentation was “unusual.” He reiterated in his supplement that “her presentation is atypical.” He also stated that although her sinus disease “may be a precipitating problem for asthma flares,” her previous surgery and sinus disease cannot be presumed to be related to her current condition. Dr. Sen opined that it would be “impossible to determine which sources of exposure contributed proportionally to [the plaintiff’s] symptoms.” He also noted that the plaintiff’s ability to perform her job would depend on her ability to perform in the work place assigned and that there are relatively few work places free of dust and mold.
The information supplied by the medical panel in the certificates, narratives, and clarifications was considered by the AM, and properly so. See Fairbairn v. Contributory Retirement Appeal Bd., 54 Mass. App. Ct. at 361-362; Houde v. Contributory Retirement Appeal Bd., 57 Mass. App. Ct. at 850 n.10. The ultimate question of causation is for CRAB. The purpose of the third question on the regional medical panel certificate is to inform CRAB whether there is a medical possibility that the events relied upon could cause the disability. The members of the panel are not prohibited from explaining their views and reasons why they reached their conclusions or from providing *135information on the medical issues involved. See Fairbairn v. Contributory Retirement Appeal Bd., supra at 361. This was not a case such as Lisbon v. Contributory Retirement Appeal Bd., 41 Mass. App. Ct. at 254-255, where the panel doctors offered their unqualified opinion on the issue of causation. Rather, the original narratives and subsequent clarifications here contained helpful information and opinion about the medical possibilities, which the AM was entitled to consider along with all the other evidence. See Fairbairn v. Contributory Retirement Appeal Bd., supra. Information and opinions contained in the narrative statements of the doctors who comprise the regional medical panel, including clarifications, may, except for unqualified opinions as to actual causation, be considered by a retirement board and CRAB on the question of causality. See ibid. As Fairbaim reminds us, the question of causation is a combination of medical and nonmedical facts. Id. at 359, quoting from Noone v. Contributory Retirement Appeal Bd., 34 Mass. App. Ct. 756, 762 (1993). The traditional use of the hypothetical question posed to the panel of doctors serves well to preserve this distinction and remind the fact finder that an expert opinion based on assumed facts can only be given weight if the assumed facts are determined to be true. See, e.g., Beaupre v. Cliff Smith & Assocs., 50 Mass. App. Ct. 480, 487 n.13 (2000). See also Liacos, Brodin, & Avery, Massachusetts Evidence § 7.7.3 (7th ed. 1999). Furthermore, even aside from reliance on expert opinion testimony under this rubric, a medical panelist may have useful information and learning to impart that will help CRAB reach an intelligent decision on issues of causation. The better the medicine or science is understood, the better a lay decision maker can make an informed judgment on such questions. On the other hand, unqualified expressions of opinion tend to mask the factual premises upon which the opinion is based.
Here, the three doctors on the panel for the most part avoided expressing unqualified opinions on causation, both in their original narratives and in their clarifications. To the extent such opinions slipped into their language, the context gave the necessary qualification. Thus, it was not improper for the AM to *136consider the medical uncertainties presented by this evidence in evaluating whether the plaintiff had carried her burden.11
The burden of proof and substantial evidence. The plaintiff bore the burden of proving causation to CRAB by a preponderance of the evidence. See Lisbon v. Contributory Retirement Appeal Bd., 41 Mass. App. Ct. at 255. On appellate review under G. L. c. 30A, § 14(7), however, we “may not set aside a CRAB decision, not otherwise legally erroneous, that is supported by substantial evidence,” Lisbon v. Contributory Retirement Appeal Bd., supra at 257, by which is meant only “ ‘such evidence as a reasonable mind might accept as adequate to support a conclusion,’ after taking into consideration opposing evidence in the record.” Ibid., quoting from G. L. c. 30A, § 1(6). Here, the plaintiff failed to persuade the AM that mold in her work place caused her permanent disability.12 In the plaintiff’s argument on appeal, there is some confusion of analysis revolving around the concept of substantial evidence in the context of fact finding in an administrative adjudicatory hearing. The plaintiff frames the issue by complaining that CRAB’s rejection of her claim on causation was not supported by substantial evidence. But it was she who was required to persuade the fact finder that her disability was proximately caused by the environment of her work place.
The plaintiff did provide the fact finder with plenty of *137evidence which could have led to a result favorable to her. If AM had been persuaded and had concluded that the plaintiff had proved her case on the issue of proximate cause, that decision would have been supported by substantial evidence in the record. But the opposing party did not have a burden to prove that the plaintiff’s condition was not caused by her work environment and to supply substantial evidence to the fact finder to support such a conclusion.
It is true that, in an adjudicatory hearing, where uncontradicted expert opinion evidence is produced on a subject beyond the common knowledge of the fact finder, there must be a basis in the record for rejecting such evidence or for remaining unpersuaded. See Robinson v. Contributory Retirement Appeal Bd., 20 Mass. App. Ct. 634, 639 (1985). See also New Boston Garden Corp. v. Assessors of Boston, 383 Mass. 456, 470 (1981) (board of assessors may not reject unrebutted and corroborated testimony “without a basis for such rejection in the record”). This rule is in place to guard against capricious and arbitrary rulings by an administrative agency. Here, as set forth above, such a basis does exist in the record.
The substantial evidence rule does not get turned on its head and applied equally both to the party that does not have the burden of proof and to the party with the burden of proof. All that is required is that the record contain facts that make rejection of such expert evidence reasonable.
The mere circumstances that a claim of causation is supported by some expert testimony does not require the AM to articulate all the reasons for rejecting that claim. Findings that disclose facts fairly detracting from the plaintiffs case are sufficient to show that the AM was not acting arbitrarily and capriciously even if those facts are not parsed and marshaled by the AM in the most convincing manner. Here, as required, the AM “made subsidiary findings of fact on the relevant issues and explained [her] reasons for rejecting the expert [evidence] based on those subsidiary findings.” Robinson v. Contributory Retirement Appeal Bd., supra at 640.
The air quality study. The plaintiff argues that the AM misread the report of Universal Engineering, the environment *138expert. Air sampling was conducted at the site of the plaintiffs office at the Ben-Hem and at her home. Mold spores were found in both locations, but in no more than moderate concentrations. For mold spores to be significant, the plaintiff needed to show that the spores to which she was allergic13 were present at her work place but not in her home. On this point the report stated as follows: “The types of growth identified in the analyses were Penicillium, Aspergillus, Cladosporium, Acremonium, Alternaría, Saprophyte, Epicoccum and Rhizopus. [The plaintiff’s] allergy testing indicated reaction to Penicillium, Aspergillus and Cladosporium.” This statement can be interpreted to mean that all the Usted types of mold were found in all samples taken during the testing. Universal Engineering knew that it was trying to shed Ught on the problem of whether offending molds could be found in both the plaintiffs work place and her home or whether they were to be found only in her work place. Thus it was not unreasonable to interpret its report as meaning that all listed molds were found in all ah samples.14 Indeed, that is the plain meaning of the language used.
In the alternative, the report could arguably be read to be no more than a Ust of all the various types of mold found at least once in a sample. This is the position argued by the plaintiff and by the dissent. The plaintiff argues that Universal Engineering must have had the data identifying the mold spores found in each sample and could have reported that information. But it is by no means obvious that the Usted types of mold do not appear commonly together in any area where any mold is to be found. The AM was entitled to reject the alternative interpretation urged by the plaintiff. The plaintiff, however, maintains that the TRB had the burden to show that the offending mold was in *139both her home and her work place. We decline to adopt this proposition.15
Judgment affirmed.
The permanency and totality of the plaintiffs disability are not at issue in this appeal, nor is the personal injury prong of the statute.
There was no live expert medical testimony at the hearing. Besides the plaintiff, the only witness who testified for the plaintiff was an expert on air quality testing. The sole witness for the TRB was the principal of the Ben-Hem.
We do not now refer to the narrative statements of the physicians on the regional medical panel, which are discussed elsewhere.
This spirometric testing only took place some time after her initial complaints. Other medical records noted the lack of objective testing.
At the hearing before the AM the plaintiffs allegation that other employees had similar complaints was challenged and unsupported except through the testimony of the plaintiff.
Except for the air quality testing report of Universal Engineering, to which the plaintiff objected, all exhibits were admitted without objection by either party. Thus, although we need not consider whether those documents were properly admitted since the issue was not raised at the hearing, Fioravanti v. State Racing Commn., 6 Mass. App. Ct. 299, 304 (1978), we conclude for the reasons stated herein that the documents were properly considered.
The certificates used here posed the following as the third question: “Is said incapacity such as might be the natural and proximate result of the personal injury sustained or hazard undergone on account of which retirement is claimed?” See G. L. c. 32, § 6(3)(a).
A copy of each request was submitted to the medical panel administrator of the Public Employee Retirement Administration Commission (PERAC). The administrator reviewed the clarifications of the panel members and found them to be in order. Letters from PERAC to the attorney for the TRB stated that the TRB was encouraged to review the clarifications and to request further clarifications if necessary, so long as PERAC was given copies.
The dissent, viewing the question posed as one calling for an unqualified opinion on causation, opines that it was improper for the question to have been placed before the panel. However, the key language of the question (“can you determine”) does not seek an unqualified opinion on causation-in-fact; rather, it probes the medical thinking that would go into the making of such a determination. In any event, for purposes of deciding whether the clarifications contained unqualified opinions on causation-in-fact that could not be given weight by the AM, in our view the important point of focus is the doctors’ responses, not the question that prompted them.
From the case law discussed supra, prohibiting the fact finder from giving weight to panelists’ unqualified opinions on causation, the dissent seems to extrapolate the principle that opinions on the medical uncertainties of establishing causation are similarly forbidden. We do not perceive support for this principle in the prior cases, and we decline to establish it now.
In her written decision, the AM explained her core reasoning as follows: “In view of the fact that the mold found in the Ben-Hem School was also found in the [plaintiff’s] home, and in view of the fact that every clinician who was asked was unable to say whether the [plaintiff’s] home or the school caused her condition, and because on the [plaintiff’s] last day of work her symptoms flared at the Lilja School where she claimed to be asymptomatic, I must conclude that the [plaintiff] has failed to meet her burden of proof with respect to causation.” The AM added, “Although the [plaintiffs] expert testifled that the air quality testing performed in 1994 probably under-reported the presence of mold because the testing was performed in the winter, and because the testing method is one that is known to kill mold, there is no other report to use. The Ben-Hem School has since been demolished. Even if one were to speculate that the prevalence of mold in the school was actually higher than that reported, the same would be true of the [plaintiff’s] home.”
Penicillium, Aspergillus, and Cladosporium.
The report also states that certain visible green mold on a floor and shower head at the Ben-Hem work place was analyzed and found to contain a type of mold to which the plaintiff was not allergic. The plaintiff criticizes this aspect of the report on the ground that it does not state that the mold mentioned was the only mold found. This was similarly a matter of reasonable interpretation of the report. Furthermore there is no discussion in the evidence about the significance of mold in a solid or viscous form to the presence of airborne mold spores.
We find no merit to the plaintiff’s contention that the AM committed an error of law by requiring proof of an “unusually traumatic event” — i.e., that the plaintiff was injured by a “traumatic release of contamination.” The AM required no such proof.