*733*550Petitioner appeals, requesting that we reverse and remand an order of the Psychiatric Security Review Board (PSRB) denying his request for discharge from PSRB jurisdiction and continuing his commitment to the Oregon State Hospital (OSH). Petitioner argues that neither PSRB's factual finding, that he presents a substantial danger to others as a result of a qualifying mental disease or defect, nor its resulting legal conclusion, that he remains subject to PSRB jurisdiction, is supported by substantial evidence. See ORS 161.351 (requiring discharge if person under PSRB jurisdiction is "no longer affected by a qualifying mental disorder or, if so affected, no longer presents a substantial danger to others that requires regular medical care, medication, supervision or treatment"). Having reviewed PSRB's order and the underlying record, we agree with petitioner. Specifically, we conclude that substantial evidence does not support PSRB's finding, in support of its order, that petitioner's anxiety disorder renders him a substantial danger to others. As a result, we reverse and remand PSRB's order asserting continuing jurisdiction over petitioner.

With the exception of PSRB's ultimate finding (that petitioner suffers from a qualifying mental disease or defect that renders him a substantial danger to others), the relevant facts are largely procedural and undisputed. Petitioner has been a patient at OSH since 1999, when a trial court entered a judgment finding petitioner guilty except for insanity and placing him under PSRB jurisdiction for a period of not more than 60 years.1 See ORS 161.327(1) (authorizing commitment to state hospital of a person found guilty except for insanity of a felony, if the person "is affected by a qualifying mental *551disorder and presents a substantial danger to others"). Since then, petitioner has had multiple hearings before PSRB. Those include one held in 2015 after OSH requested that PSRB discharge petitioner because he no longer carried a "jurisdictional diagnosis," i.e. , a qualifying mental health diagnosis warranting continued PSRB jurisdiction. See ORS 161.341(1) (2011), amended by Or. Laws 2017, ch. 634, § 11, and Or. Laws 2017, ch. 442, § 4;2 OAR 859-060-0040. OSH recognized that petitioner remained at risk of sexually reoffending due to a nonqualifying diagnosis *734of Pedophilic Disorder.3 OSH also acknowledged that petitioner exhibited symptoms of anxiety. In OSH's view, however, petitioner's anxiety symptoms were associated with his pedophilia and another nonqualifying personality disorder. In any event, OSH maintained, those symptoms were not themselves causally related to petitioner's dangerousness. Further, by that time, petitioner's only other potentially jurisdictional diagnosis of Autism Spectrum Disorder could no longer be supported. Thus, OSH asserted, PSRB no longer had any basis on which to continue exercising jurisdiction over *552petitioner. At the conclusion of that proceeding, however, PSRB concluded that, because petitioner was "affected by a mental disease or defect which, when active, render[ed] him a substantial danger to others," he was properly under PSRB jurisdiction; accordingly, PSRB denied the requested discharge and continued petitioner's commitment to OSH. Petitioner appealed that decision.

In 2016, while that appeal of PSRB's 2015 decision was pending before us, petitioner initiated another PSRB hearing on his own behalf. OSH again supported petitioner's discharge from PSRB jurisdiction. By then, however, OSH no longer held the view that petitioner did not have a potentially jurisdictional diagnosis; rather, acknowledging that petitioner had current diagnoses for Illness Anxiety Disorder and Other Specified Anxiety Disorder, OSH took the position that neither of those otherwise qualifying diagnoses was the cause of petitioner's dangerousness. As before, OSH expressed its view that the substantial risk that petitioner would engage in pedophilic behaviors in the future was solely due to his pedophilia. Accordingly, OSH again maintained that PSRB lacked jurisdiction over petitioner. The present appeal arises from PSRB's February 2017 hearing, at which it again rejected that argument.

To provide context for our discussion of the evidence presented at that hearing, we pause to give an overview of the procedural and substantive laws governing PSRB's consideration of petitioner's request. At the time of the hearing, petitioner's request was governed by ORS 161.351 (2011), amended by Or. Laws 2017, ch. 634, § 13, and Or. Laws 2017, ch. 442, § 5. Under that provision, the standards for discharge were as follows:

"(1) Any person placed under the jurisdiction of the Psychiatric Security Review Board or the Oregon Health Authority under ORS 161.315 to 161.351 shall be discharged at such time as the agency having jurisdiction over the person, upon a hearing, finds by a preponderance of the evidence that the person is no longer affected by mental disease or defect or, if so affected, no longer presents a substantial danger to others that requires regular medical care, medication, supervision or treatment.

*553"(2) For purposes of ORS 161.315 to 161.351, a person affected by a mental disease or defect in a state of remission is considered to have a mental disease or defect. A person whose mental disease or defect may, with reasonable medical probability, occasionally become active and when it becomes active will render the person a danger to others may not be discharged. The person shall continue under supervision and treatment necessary to protect the person and others.

"(3) In determining whether a person should be committed to a state hospital or secure intensive community inpatient facility, conditionally released or discharged, the board and the authority shall have as their primary concern the protection of society."

As we observed in Beiswenger v. PSRB , 192 Or. App. 38, 41, 84 P.3d 180, rev. dismissed as improvidently allowed , 337 Or. 669, 104 P.3d 601 (2004), the legislature did not define "mental disease or defect" other than to say what those terms do not mean, specifically, that they "do not include an abnormality manifested only by repeated criminal or otherwise antisocial conduct, nor do they include any abnormality constituting solely a personality disorder."

*735ORS 161.295(2) (1983), amended by Or. Laws 2017, ch. 634, § 3. Likewise, the legislature had not defined "personality disorder." See Beiswenger , 192 Or. App. at 45, 84 P.3d 180. Based, however, on our review of the statutory text and context, as well as the available legislative history and applicable canons of construction, we concluded in Beiswenger that, as used in ORS 161.295(2) (1983), "personality disorder" included sexual conduct disorders, alcohol dependency, and drug dependency. Id . at 54, 84 P.3d 180. As a result, those disorders did not constitute "mental disease[s] or defect[s]" within the meaning of ORS 161.295(2) (1983). Id . And, in petitioner's case, there is no dispute that pedophilia is a sexual conduct disorder and therefore not a qualifying mental disease or defect for purposes of PSRB jurisdiction. Thus, under ORS 161.351(1) (2011), PSRB was required to discharge petitioner from its jurisdiction unless it found, by a preponderance of the evidence, that petitioner was affected by a mental health condition other than pedophilia or another "personality disorder" and that he "present[ed] a substantial danger to others that require[d] regular medical *554care, medication, supervision or treatment." See also OAR 859-050-0055.4

We return to the facts. At the 2017 hearing, PSRB considered petitioner's voluminous OSH file, which documented his various (and varying) mental health diagnoses, OSH's efforts to provide treatment in connection with those diagnoses, and petitioner's progress in that regard. Petitioner's records show that, well before he was first hospitalized in 1999, petitioner had been diagnosed with autism, a qualifying mental disease or defect. It is not readily apparent, however, whether that diagnosis played any role in PSRB's initial assertion of jurisdiction, because the judgment finding petitioner guilty except for insanity does not reflect a specific diagnosis. See ORS 161.325 (requiring qualifying mental disorder to be identified on the record but not explicitly requiring its listing in the judgment). In any event, by the time of petitioner's PSRB hearing in 2015, his treatment providers all agreed that an autism diagnosis could no longer be supported. And, although petitioner appears also to have been diagnosed shortly before his initial commitment with schizophrenia, another qualifying mental disease or defect, the record does not reflect an ongoing diagnosis for that or any related disorder, nor does PSRB appear to have relied on such a diagnosis in either 2015 or 2017.5 Notably, since at least 2006, petitioner has openly acknowledged having malingered at the time of his *555retrial as to certain "psychotic" symptoms in order to be hospitalized rather than incarcerated.

PSRB also heard testimony, including that of Dr. Roff, who at the time of the hearing had been petitioner's treating psychiatrist for approximately two years. Roff, who, as part of her evaluation had reviewed the thousands of pages of petitioner's OSH records, testified that she had diagnosed petitioner with the following disorders, all of which are recognized in the Diagnostic and Statistical Manual of Mental Disorders-Fifth Edition (DSM-5): Illness Anxiety Disorder (previously known as Hypochondriasis ); Other Specified Anxiety Disorder; Pedophilic Disorder (identified by Roff as petitioner's primary diagnosis); Antisocial Personality Disorder ; Obsessive-Compulsive Personality Disorder ; and Alcohol Use Disorder in full remission in *736a controlled environment. The parties evidently agree that, of those six diagnoses, only two qualify as "mental disease[s] or defect[s]" for purposes of PSRB jurisdiction: Illness Anxiety Disorder and Other Specified Anxiety Disorder. See ORS 161.295(2) (1983) ("mental disease or defect" does not include personality disorders); Beiswenger , 192 Or. App. at 54, 84 P.3d 180 (recognizing paraphilia to be a sexual conduct disorder and therefore a nonqualifying "personality disorder" under ORS 161.295(2) (1983) ).6 Consistent with OSH's view, however, Roff testified that petitioner's dangerousness was due solely to his nonqualifying diagnosis of Pedophilic Disorder.

Following the 2017 hearing, PSRB made the following findings relevant to this appeal, accompanied by references to exhibits in the record that PSRB relied upon for support:

"2. [Petitioner] is affected by a mental disease or defect * * * [and] continues to suffer from an anxiety disorder, to *556wit, Illness Anxiety Disorder and Other Specified Anxiety Disorder, for which he currently receives medication. ***

"The Board carefully considered the concluding opinion of Dr. Roff that [petitioner's] dangerousness is solely caused by his non-qualifying mental conditions, however, the Board is not convinced by a preponderance of the evidence of her conclusion. Specifically, the Board notes Dr. Roff's testimony and progress notes allow for the inference that [petitioner's] anxiety disorder makes him less likely to seek out psychiatric treatment, including sex offender treatment, which makes him a substantial danger to others.

"3. [Petitioner], without adequate supervision and treatment, would continue to present a substantial danger to others ***. ***

"The Board additionally notes that when asked directly about [petitioner's] risk of dangerousness now versus at the time of his last hearing, Dr. Roff testified that [petitioner's] risk of dangerousness to others has increased, in her opinion due to learning more about [petitioner] and his propensities to engage in high-risk behavior.

"4. The State sustained its burden of proving by a preponderance of the evidence that [petitioner] continues to be affected by a mental disease or defect and continues to be a substantial danger to others and that he should not be discharged from the jurisdiction of the Board.

"5. [Petitioner] could not be adequately controlled and treated in the community if he were conditionally released at this time."

Based on those findings, PSRB concluded as a matter of law that:

"1. [Petitioner], being affected by a mental disease or defect which, when active, renders him a substantial danger to others, is under the jurisdiction of the Psychiatric Security Review Board.

"2. [Petitioner] is not a proper subject for conditional release because he could not be adequately controlled and treated in the community and therefore it would not be in the best interest of justice and the protection of society to release him at this time."

*557Accordingly, PSRB ordered that petitioner's commitment to a state hospital be continued under ORS 161.346(1)(c) and ORS 161.351(2). Petitioner appeals from that order.

We review PSRB's order for substantial evidence. Specifically, ORS 183.482(8)(c) provides that a reviewing court

"shall set aside or remand the order if the court finds that the order is not supported by substantial evidence in the record. Substantial evidence exists to support a finding *737of fact when the record, viewed as a whole, would permit a reasonable person to make that finding."

We review both PSRB's factual findings and its legal conclusions derived from those findings. Knotts v. PSRB , 250 Or. App. 448, 454, 280 P.3d 1030 (2012). We review any disputed findings to determine whether they are supported by the evidentiary record and, specifically, whether a reasonable person, viewing the record as a whole, could make those findings. Id. at 454-55, 280 P.3d 1030. We do not substitute our own findings for those made by PSRB. Rather, we determine whether PSRB could reasonably make the findings that it made. Id. at 455, 280 P.3d 1030. In reviewing whether substantial evidence supports PSRB's findings, we " 'evaluate the substantiality of supporting evidence by considering all the evidence in the record.' " Garcia v. SAIF , 187 Or. App. 51, 57, 66 P.3d 522 (2003) (quoting Younger v. City of Portland , 305 Or. 346, 356, 752 P.2d 262 (1988) (emphasis in Garcia )).

In turn, we review PSRB's legal conclusions to determine whether they logically follow from the facts PSRB found. Knotts , 250 Or. App. at 455, 280 P.3d 1030. In doing so, we consider only the reasoning stated in the challenged order. Id. ; see also Castro v. Board of Parole , 232 Or. App. 75, 85-86, 220 P.3d 772 (2009) ("Our duty is to evaluate the board's logic, not to supply it."). PSRB must, therefore, explain its conclusions in sufficient detail to allow us to review its reasoning. If PSRB's reasoning is not apparent from the order or if that reasoning is faulty, the order is not supported by substantial reason. Knotts , 250 Or. App. at 455, 280 P.3d 1030.

On appeal, petitioner does not appear to dispute either that he has a qualifying mental disease or defect or that he is a substantial danger to others. Rather, he argues *558that substantial evidence does not support PSRB's finding that he is dangerous because of a qualifying mental disease or defect. Petitioner's dangerousness, he contends, results from his Pedophilic Disorder, which, as noted, is not a mental disease or defect for purposes of PSRB jurisdiction. Thus, according to petitioner, PSRB erred in finding that he presents a substantial danger to others as a result of a qualifying mental disease or defect.

In response, PSRB argues that the record supports the finding that petitioner's Other Specified Anxiety Disorder makes him "more dangerous than he otherwise would be," thereby rendering him a substantial danger to others and warranting continued PSRB jurisdiction. Specifically, PSRB contends, petitioner's Other Specified Anxiety Disorder makes it more likely that, if petitioner were to be discharged from PSRB jurisdiction, he would forgo necessary sex offender treatment, seek out settings in which it would be difficult to detect his predatory behavior, and carefully plan out and execute his pedophilic activities so as to avoid apprehension. Thus, even though PSRB does not dispute that petitioner's Pedophilic Disorder is the primary cause of his dangerousness, PSRB reasons that the Other Specified Anxiety Disorder makes it more likely than it would be in the absence of that disorder that petitioner will reoffend; that is, because the danger brought on by petitioner's pedophilia will, due to petitioner's anxiety, not be lessened , his Other Specified Anxiety Disorder makes him more dangerous than he would be if he did not have that particular mental disease or defect. That, PSRB argues, is all that ORS 161.351(1) (2011) requires to warrant continued jurisdiction.

As the foregoing arguments demonstrate, the focus of both parties is whether there is a causal connection-or "nexus"-between petitioner's qualifying mental disease or defect and the dangerousness he poses. In exploring that issue, we note that, although neither the plain text of ORS 161.351(1) (2011) nor any case law of which we are aware establishes a nexus requirement, neither party advances an alternate reading of the statute. Moreover, even though some of the language in PSRB's order can be viewed as *559suggesting that only a temporal-as opposed to causal-relationship is required, the board's ultimate conclusion is stated in terms of ORS 161.351(2) (2011), the plain text of which does suggest a causality requirement by considering whether a person's mental disorder "will render the person a danger to others." (Emphasis *738added.) And, because that is the ruling that petitioner challenges, we, like the parties, proceed to consider whether substantial evidence and reason support PSRB's finding of a causal relationship between any qualifying mental disease or defect that petitioner has and the substantial danger to others that he presents.

As noted, PSRB ultimately concluded that petitioner, "being affected by a mental disease or defect which, when active, renders him a substantial danger to others, is under the jurisdiction of the Psychiatric Security Review Board." (Emphasis added.) PSRB explained its underlying finding of a causal relationship between petitioner's qualifying mental disease or defect and his dangerousness to others as follows:

"[Petitioner] is affected by a mental disease or defect *** [and] continues to suffer from an anxiety disorder, to wit, Illness Anxiety Disorder and Other Specified Anxiety Disorder, for which he currently receives medication. ***

"The Board carefully considered the concluding opinion of Dr. Roff that [petitioner's] dangerousness is solely caused by his non-qualifying mental conditions, however, the Board is not convinced by a preponderance of the evidence of her conclusion. Specifically, the Board notes Dr. Roff's testimony and progress notes allow for the inference that [petitioner's] anxiety disorder makes him less likely to seek out psychiatric treatment, including sex offender treatment, which makes him a substantial danger to others."7

(Emphasis added.)

*560Having reviewed the record, we agree with petitioner that PSRB's finding of that nexus is not supported by substantial evidence, because the record, viewed as a whole, would not permit a reasonable person to make that finding. See ORS 183.482(8)(c). We focus on petitioner's Other Specified Anxiety Disorder, because PSRB did not find and does not argue on appeal that his Illness Anxiety Disorder bears any relationship to petitioner's dangerousness. Roff testified in detail at the 2017 hearing about petitioner's Other Specified Anxiety Disorder diagnosis. She explained that the diagnosis applies to individuals who exhibit significant symptoms characteristic of an anxiety disorder, but who do not meet the criteria for any full DSM-5 diagnosis; the diagnosis is appropriate when a treatment provider observes "something more" in an individual's behaviors than one would expect to see in a typical individual. Roff explained that petitioner's anxiety centers around the fact that he is known both in the hospital and in the community as a sex offender. The primary source of petitioner's anxiety is his fear of retaliation, whether retaliation that he may suffer in the future or that he perceives himself as having narrowly avoided in the past. For example, petitioner is certain that he would have been killed if he had not been able to overturn his prison sentence. According to Roff, however, that aspect of petitioner's anxiety is situational and normative, meaning that, despite the severity of petitioner's anxiety, it is not out of proportion to his reality; that is, to the threats he actually faces.

In finding that petitioner's Other Specified Anxiety Disorder renders him a substantial danger to others, PSRB expressly rejected Roff's opinion that petitioner's dangerousness resulted solely from his nonqualifying diagnosis of pedophilia, and therefore not from his qualifying Other Specified Anxiety Disorder. PSRB reasoned that "Roff's testimony and progress notes allow for the inference that [petitioner's] anxiety disorder makes him less likely to seek out psychiatric treatment, including sex offender treatment, which makes him a substantial danger to others." Because PSRB expressly rejected Roff's opinion based on its contrary findings of fact, we review whether substantial evidence supports those findings. Stated differently, "[a]n assertion of *561a finding of fact as part of an explanation for disregarding evidence is subject to attack if that fact *739relied upon is not, itself, supported by substantial evidence." Garcia v. Boise Cascade Corp. , 309 Or. 292, 296, 787 P.2d 884 (1990).

Viewed in that light, we conclude that the record lacks substantial evidence to support PSRB's finding, purportedly based on Roff's testimony and progress notes, that petitioner's "anxiety disorder makes him less likely to seek out psychiatric treatment, including sex offender treatment, which makes him a substantial danger to others. " (Emphasis added.) Roff's overall sense, based largely on her clinical experience with petitioner and her review of his extensive records of treatment with other providers, was that there was no risk that petitioner's anxiety could drive him toward sexual recidivism. In support of that opinion, Roff also cited a scientific meta-study that she had reviewed and that had examined whether a host of variables were correlated in any way with a risk of sexually reoffending. Roff explained, without objection, that the meta-study found no statistically significant correlation between anxiety and the risk for sexual recidivism.

Roff further explained that, in her opinion, petitioner's anxiety is primarily inhibitory; that is, it would tend to make it less likely for him to sexually reoffend. Roff acknowledged that petitioner's anxiety could lead him to leave the country and thus contribute to certain preoffense behaviors. And, when asked whether petitioner's anxiety would "prevent him or have any kind of impact on whether or not he participates" in sex offender treatment, Roff initially agreed that it could. Similarly, she seemingly accepted that petitioner's anxiety disorder "could influence whether or not he remains safe because it's influencing his ability to get treatment[.]" Roff later testified, however, that petitioner had stated "very clearly that he has no issue whatsoever" in participating in sex offender treatment outside the hospital and is "absolutely willing" to restart treatment in the community. Thus, Roff clarified, "if your real question is if you let [petitioner] out the door is the anxiety going to get in the way of him participating in a sex offender treatment program, I would say that the evidence points against that."

*562Given Roff's express testimony that petitioner's anxiety would not interfere with his participation in sex offender treatment, no reasonable person would rely on her testimony and the underlying progress notes-as PSRB purported to do-to draw the opposite inference, i.e. , that petitioner's "anxiety disorder makes him less likely to seek out psychiatric treatment, including sex offender treatment[.]"8 See ORS 183.482(8)(c) (substantial evidence is such evidence as would allow a reasonable person to make the finding in question). Thus, unless (and despite PSRB's apparent reliance only on Roff's testimony and progress notes) other evidence in the record supports that inference, PSRB erred in making that finding. See id . (requiring record to be "viewed as a whole").

Turning to the balance of the record, we conclude that it does not support PSRB's rejection of Roff's opinion and concomitant substitution of its own finding that petitioner's anxiety will prevent him from participating in treatment. See Garcia , 309 Or. at 296, 787 P.2d 884 (requiring that an agency's assertion of a factual finding as a basis to reject evidence be supported by substantial evidence). Notably, while petitioner resided at OSH from 1999 to 2008, as well as for some time after he returned to OSH in 2013, he participated in the sex offender treatment program offered at that facility.9 And, when petitioner was on a conditional release from OSH, he actively engaged in treatment, albeit while also demonstrating several highly concerning behaviors that ultimately led to the revocation of his conditional release. PSRB has not identified, nor have we discovered in our review, any evidence in the record that can support the inference that petitioner's anxiety will impede his participation in treatment.

*740Moreover, the record does not support the balance of PSRB's finding, namely, that petitioner's failure to participate *563in treatment "makes him a substantial danger to others." In arguing to us on appeal that petitioner's anxiety-and its ostensible consequence that he will not get the treatment he needs-"makes him even more dangerous than he otherwise would be," PSRB appears to argue that petitioner's anxiety disorder "renders" him dangerous by not lessening the danger already presented by his pedophilia. Whatever merit that argument might otherwise have, we must reject it here. As noted, PSRB offers that argument for the first time in its briefing on appeal. PSRB's order did not, however, provide any rationale-much less that one-for its finding that petitioner's failure to participate in treatment "makes him a substantial danger to others." And, it is PSRB's obligation to provide its reasoning in its order; we will not speculate as to what PSRB's reasoning might have been, nor can we rely on reasoning PSRB might belatedly provide in its briefing on appeal. See Castro , 232 Or. App. at 85-86, 220 P.3d 772 ("Our duty is to evaluate the board's logic, not supply it."); Knotts , 250 Or. App. at 455, 280 P.3d 1030 (noting that our consideration is limited to the reasoning stated in the challenged order). As a result, the finding that petitioner's failure to participate in treatment "makes him a substantial danger to others" is not supported by substantial evidence in the record. Knotts , 250 Or. App. at 455, 280 P.3d 1030 (substantial evidence standard requires that an order reveal the logic that led an agency to draw the inferences that it did from the evidence it considered).

As for petitioner's potential to engage in preoffense behaviors, we recognize that Roff's May 2015 patient progress notes did identify that one "significant connection" between petitioner's anxiety and his risk of reoffending. Specifically, Roff noted petitioner's fear, discussed above, of retaliation from others due to his status as a sex offender. Roff believed that petitioner would likely leave the State of Oregon and possibly the United States altogether to avoid that threat. Roff further believed that, once outside of the United States, petitioner's access to potential victims would be greater and the perceived threat of reincarceration-which otherwise would tend to inhibit reoffending-would decrease. As Roff summarized,

"[petitioner's] anxiety over his future safety makes it likely that he will move abroad, where he will have access to a *564pool of vulnerable victims and where he is likely to reoffend. However, while his anxiety is indirectly related to his future risk of sexual offending, I cannot state that it is a direct cause."

To the extent that PSRB contends that this testimony supports the inference that petitioner's anxiety makes him more dangerous than his unregulated pedophilia acting alone, we find two difficulties with that argument. The most obvious difficulty is that the challenged order bears no indication that PSRB perceived the evidence in that manner; as a result, that inference would be unsupported for the same reason as the inference regarding petitioner's participation in treatment. The other difficulty is that, even if, in isolation, Roff's testimony-that petitioner's anxiety would not be the "direct cause" of sexual reoffending-can support the inference that it nonetheless would be a cause of that danger, it is, at least in the absence of explanation from PSRB, too thin a reed to bear that weight when viewed in light of the record as a whole.

Specifically, that understanding of Roff's testimony fails to grapple with, among other things, the view of OSH's risk review panel. See Younger , 305 Or. at 354, 752 P.2d 262 ("The substantiality of evidence must take into account whatever in the record fairly detracts from its weight." (Internal quotation marks omitted.)). The risk review panel, which relied substantially on the recommendation of Roff and petitioner's interdisciplinary treatment team, was itself composed of mental health professionals, including a forensic risk psychologist, a forensic social worker, a forensic supervising psychiatrist, and an Oregon Health Authority PSRB liaison. On behalf of OSH, the risk review panel supported petitioner's jurisdictional discharge based, in part, on its finding that,

"[a]lthough [petitioner] has jurisdictional diagnoses of Illness Anxiety Disorder and *741Other Specified Anxiety Disorder, these diagnoses and resultant symptoms do not seem to influence his future dangerousness and engaging in future pedophilic behaviors. Anxiety symptoms generally inhibit sexual desire and activity. Anxiety can occasionally be relieved by compulsive masturbatory behaviors [such as petitioner had exhibited], but that is a very *565different activity from engaging in predatory pedophilic behaviors."

The risk review panel recognized that petitioner continues to be "at elevated risk for engaging in deliberate, planned and targeted pedophilic behaviors when unsupervised in the community," but the panel clarified that "this risk is solely due to his diagnosis of a Pedophilic Disorder and not due to any other jurisdictional diagnosis." (Emphasis added.) Rather than consider Roff's testimony in light of the risk review panel's assessment, PSRB appears to have disregarded the panel's view entirely.

PSRB's finding further disregards other testimony presented at the hearing. Dr. Johnson-a forensic psychologist-also testified on behalf of petitioner. Like Roff, he testified that petitioner is "undoubtedly" at risk to commit further sexual offenses, but that risk is not causally related to anything other than petitioner's Pedophilic Disorder. When asked about petitioner's anxiety, Johnson stated:

"Anxiety is part of the human condition. We all have anxiety. Anxiety ebbs and flows for everybody on the face of the earth. [Petitioner] is no different. He has certainly experienced an exacerbation of his anxiety because of his offense history and interaction with the criminal justice system. He has good grounds to be anxious. But once again it did not cause him to develop pedophilia and it does not drive his pedophilia. So to the extent that it is interactive at all, the hope, the expectation is that it will be a reminder if and when he begins to engage in slippery thinking or slippery behavior that could lead to an offense. *** So in that sense I would agree that it stands a very good chance of being inhibitory. I can think of no arguments to suggest that it actually promotes or encourages him to sexually act out."

When asked what role of cause and effect petitioner's anxiety has for his dangerousness in the community, Johnson replied, "none whatsoever." Johnson further testified that, in his opinion, petitioner's response to treatment and willingness to participate in treatment is independent of his anxiety. Specifically, when asked whether the fact that petitioner has anxiety would influence whether or not he will go *566to treatment, Johnson indicated that it would not. As with the opinion of the risk review panel, there is no indication in PSRB's order that it gave Johnson's testimony any consideration at all, nor that there was substantial evidence in the record to support PSRB's rejection of that testimony.

In recounting the evidence that PSRB disregarded without explanation in favor of its own opinion, we note that this is not a case where PSRB simply found one doctor's opinion more persuasive than another. See Minor v. SAIF , 290 Or. App. 537, 551, 415 P.3d 1107 (2018) (citing Armstrong v. Asten-Hill Co. , 90 Or. App. 200, 206, 752 P.2d 312 (1988) ("[I]f there are doctors on both sides of a medical issue, whichever way the Board finds the facts will probably have substantial evidentiary support.")). Rather, in this case, PSRB, without explanation, (a) formed its own opinion that petitioner's anxiety would impede his participation in sex offender treatment, something that no doctor (or any other witness) endorsed; (b) implicitly found, based upon Roff's testimony or other unidentified evidence in the record that petitioner's anxiety would cause him to travel to other countries where he would present an increased threat to others; or (c) both of those things. And if, in fact, either rationale is the basis of PSRB's finding that the nexus at issue exists, it was required to disclose that reasoning and the facts to support it in its order, which it did not do. As a result, we conclude that PSRB's order is not supported by substantial evidence.

In concluding that substantial evidence does not support PSRB's finding-that petitioner suffers from a qualifying mental disease or defect that renders him a substantial *742danger to others-we are not suggesting that a qualifying diagnosis, such as petitioner's anxiety disorder, could never combine with a nonqualifying diagnosis, such as his pedophilia, so as to warrant PSRB jurisdiction. Rather, we are simply not persuaded that, in this case, substantial evidence in this record supports PSRB's finding that petitioner's Other Specified Anxiety Disorder in any way caused him to be dangerous in a manner that his Pedophilic Disorder did not already cause. Without that evidence, neither that finding, nor the resultant conclusion that *567petitioner remains subject to PSRB's jurisdiction, can stand. Accordingly, we reverse and remand the challenged order. See ORS 183.482(8)(c) ("The court shall set aside or remand the order if the court finds that the order is not supported by substantial evidence in the record."); cf. Drew v. PSRB , 322 Or. 491, 499, 909 P.2d 1211 (1996) (remanding order in case in which "the record contain[ed] substantial evidence that could support PSRB's finding that petitioner was a substantial danger to others, but PSRB did not connect its decision to that evidence").

Reversed and remanded.