ORDER AND JUDGMENT*
This case requires us to answer two questions: First, whether a Colorado statute operates retroactively, where it affects the standard of review courts use to interpret plans governed by the Employee Retirement Income Security Act (ERISA). Second, whether the insurer in this case abused its discretion in denying a claimant disability benefits under the parties’ governing plan. To both questions, the district court answered no and we agree.
I
As an employee of the Kroger Company, Mary McClenahan was entitled to benefits under the company’s health and welfare benefits plan. As part of this ERISA plan, Metropolitan Life Insurance Company provided long-term disability insurance to Kroger employees. Due to a neuromusculoskeletal condition, radiculopathy, Ms. McClenahan eventually stopped working for Kroger, and MetLife provided her benefits for a twenty-four month period that expired March 13, 2006.
After that period, MetLife terminated payments to Ms. McClenahan. Pursuant to a limitation clause in the plan, MetLife determined that Ms. McClenahan couldn’t *695receive benefits past a twenty-four month period unless she demonstrated continuing “objective evidence of ... radiculopathies” — radiculopathies defined in the plan as a “[djisease of the peripheral nerve roots supported by objective clinical findings of nerve pathology.” Aplt-App. Vol. Ill at 385. In making its decision to deny Ms. McClenahan continued benefits past the initial twenty-four month period, Met-Life relied in part on Dr. R. Kevin Smith’s independent review of her medical file and his conclusion that evidence of radiculopathy was lacking.
Ms. McClenahan challenged this decision through MetLife’s internal appeal procedures, disputing MetLife’s factual assessment that she could no longer show objective evidence of radiculopathies. Ms. McClenahan forwarded to MetLife an electromyography test conducted on March 9, 2006 by Dr. John Stephens, who reported that “there is some EMG evidence of what appears to be chronic denervation likely in a left L5 pattern.” Aplt. App. Vol. Ill at 564. For its part, during the claims review process MetLife consulted other in-house medical advisors and independent physicians besides Dr. Smith, including Dr. Joseph Monkofsky who concluded that neither the March 9 EMG report nor other medical files demonstrated “any of the exclusionary diagnoses listed ... including current objective evidence of radiculopathy.” ApltApp. Vol. Ill at 528. MetLife also asked Ms. McClenahan’s physician, Dr. John Drye, to review MetLife’s findings. Disagreeing with the other physicians consulted by MetLife, Dr. Drye believed Ms. McClenahan’s disability had objective evidence of radiculopathies. ApltApp. Vol. Ill at 509. Based on the entire record before it, Met-Life nevertheless upheld its original determination to deny Ms. McClenahan any further benefits.
Having exhausted MetLife’s appeals process, Ms. McClenahan brought this lawsuit, arguing that MetLife erred in its assessment of her condition and arguing that she should be allowed to supplement the administrative record with a new medical report dated March 19, 2008. The district court held the new medical report inadmissible and eventually granted summary judgment in favor of MetLife.
II
In this appeal, Ms. McClenahan contends that a newly enacted Colorado statute subjects MetLife’s claims decisions to de novo review, and that the district court erred by applying a lesser, abuse of discretion standard to MetLife’s determination. Further, even if an abuse of discretion standard does apply, Ms. McClenahan argues that MetLife abused its discretion. Finally, Ms. McClenahan contests the district court’s exclusion of the March 19, 2008 medical report. We address each of these arguments in turn.
A
The United States Supreme Court has told us that the denial of benefits under an ERISA plan is reviewed for abuse of discretion if the plan gives the plan fiduciary — here, MetLife — discretionary authority to determine eligibility for benefits or to construe the terms of the plan. See Firestone Tire & Rubber Co. v. Bruch, 489 U.S. 101, 115, 109 S.Ct. 948, 103 L.Ed.2d 80 (1989). Before us both sides agree that the policy gives MetLife discretion, and the district court applied the abuse of discretion standard when conducting its review of MetLife’s disability determination.
Still, Ms. McClenahan argues the district court erred when it reviewed Met-Life’s decision for an abuse of discretion; in her view, the court should have applied *696 de novo standard of review. This is so, she insists, because of a relatively recent Colorado statute requiring “de novo [review] in any court with jurisdiction” for a disability benefits “claim [that] has been denied in whole or in part.” Colo.Rev. Stat. § 10-3-1116(3). Notably, neither side before us suggests that Colorado’s statute is preempted by ERISA. Instead, they disagree only over the application of the state statute to this suit according to its own terms. Because we agree with MetLife that the statute doesn’t apply to this suit we have no occasion to pass on the question whether, if applicable, the Colorado statute might or might not be preempted by ERISA.
To apply to this case at all, the Colorado statute must operate retroactively. That’s because § 10-3-1116 was enacted after all the events at issue had occurred, including the plan’s formation and MetLife’s denial of benefits. See In re Estate of DeWitt, 54 P.3d 849, 854 (Colo.2002) (“[A] statute is presumed to operate prospectively, meaning it operates on transactions occurring after its effective date.”) (emphasis added). In deciding whether a statute can be applied retroactively under Colorado law, a court must ask two questions. First, the court must ask if the Colorado General Assembly clearly intended the statute to have retroactive effect. Second, if the legislature did so intend, the court must then ask if the statute is unconstitutionally retrospective. See DeWitt, 54 P.3d at 854; Colo. Const. Art. II, § 11 (prohibiting any statute that is “retrospective in its operation.”) (emphasis added). A statute intended to operate retroactively is impermissibly retrospective under the Colorado constitution — and thus can’t be applied retroactively under state law — if but only if it impairs a vested right, creates a new obligation, imposes a new duty, or attaches a new disability. See DeWitt, 54 P.3d at 854; Cont’l Title Co. v. District Court, 645 P.2d 1310, 1314-15 (Colo.1982) (allowing retroactive application of a law where its only effect was to “provide an alternative remedy for vindication of the alleged discriminatory and unfair employment practice”).
Conflating the terms “retroactive” and “retrospective,” Ms. McClenahan argues that Colo.Rev.Stat. § 10-3-1116(3) can apply to this case because it doesn’t constitute an impairment of a vested right or imposition of a new duty. After all, she says, the statute only affects a court’s standard of review. But even assuming without deciding that the statute before us doesn’t impair any vested right and so isn’t retrospective, Ms. McClenahan fails to show that the legislature clearly intended the statute to have retroactive effect — the first step of the two-part test under Colorado law. Indeed, at oral argument Ms. McClenahan’s counsel admitted he couldn’t discern any such intent, and neither do we. While an express “declaration” of retroactivity isn’t required, there is no indication that the statute before us was ever intended to apply to prior conduct. Cf. Ficarra v. Dep’t of Regulatory Agencies, 849 P.2d 6, 12-14 (Colo.1993) (finding “unmistakable intent” of retroactivity, where the statute on its face would ascribe “to certain transactions that occurred before the effective date of the [amendment] different legal effects from that which they had under the law when they occurred.”). So it is that under Colorado Supreme Court precedent, the statute here can’t be applied to Met-Life’s claim determination, and the district court properly reviewed MetLife’s decision in this case for abuse of discretion.
B
We now turn to the propriety of the district court’s holding that MetLife did not abuse its discretion, or act arbi*697trarily and capriciously, when denying Ms. McClenahan’s claim for continued benefits. In the ERISA context, the terms “abuse of discretion” and “arbitrary and capricious” are used interchangeably, at least in this circuit. Weber v. GE Group Life Assur. Co., 541 F.3d 1002, 1010 n. 10 (10th Cir. 2008) . And a decision is arbitrary and capricious if it lacks a “reasonable basis” or is not supported by substantial evidence. See Adamson v. Unum Life Ins. Co. of Am., 455 F.3d 1209, 1212 (10th Cir.2006) (“[T]here is no requirement that the basis be the only logical one or even the superlative one ... [only that] the administrator’s decision resides somewhere on a continuum of reasonableness— even if on the low end.”); Sandoval v. Aetna Life & Cas. Ins. Co., 967 F.2d 377, 382 (10th Cir.1992). Based on the record as a whole, evidence is substantial if a “reasonable mind might accept [it] as adequate to support the conclusion reached by the [decisionmaker].” Sandoval, 967 F.2d at 382 (citation omitted). While the decisionmaker must take into account “whatever in the record fairly detracts from” its determination, substantial evidence requires “more than a scintilla but less than a preponderance.” Id. (citation omitted).
Because MetLife suffers from a conflict of interest in this case — a conflict created by the fact that MetLife pays benefits out of its own pocket and determines whether an employee is eligible — we “dial back our deference,” weighing the conflict as a “factor in determining whether there is an abuse of discretion.” Weber, 541 F.3d at 1010; Holcomb v. Unum Life Ins. Co. of Am., 578 F.3d 1187, 1192-93 (10th Cir. 2009) . The plaintiff nevertheless retains the burden of showing abuse. See Metro. Life Ins. Co. v. Glenn, 554 U.S. 105, 116, 128 S.Ct. 2343, 171 L.Ed.2d 299 (2008); Holcomb, 578 F.3d at 1192-93 (10th Cir. 2009) (recognizing abrogation of Fought v. UNUM Life Ins. Co. of Am., 379 F.3d 997, 1005 (10th Cir.2004)).
We agree with the district court that Ms. McClenahan failed to carry her burden in this case. This is because we can’t say MetLife acted unreasonably when it denied benefits after she failed to provide evidence of radiculopathies on or after March 13, 2006. MetLife consulted in-house nurses and doctors, as well as independent physicians, and sought out responses from Ms. McClenahan’s own treating physicians. Cf. Holcomb, 578 F.3d at 1193 (“Unum took steps to reduce its inherent bias by hiring two independent physicians----”). For example, independent physician Dr. Smith concluded that “[a]s it relates to the claimant’s neuromusculoskeletal and soft tissue disorders, the medical records do not indicate objective clinical evidence for ... radiculopathies....” Aplt.App. Vol. IV at 636. Moreover, Dr. Smith wrote that “[r]ecent exam findings” indicated “no neuromuscular abnormalities or nerve root tension signs consistent with radiculopathies.” Id. After Ms. McClenahan challenged Met-Life’s initial claim denial, MetLife consulted another independent physician, Dr. Monkofsky. While asked by MetLife to look for evidence of radiculopathies “beyond March 13, 2006,” Dr. Monkofsky conducted a more extensive review, examining Ms. McClenahan’s records dating from October 2003 through October 2007 — the date he performed his review.1 See Aplt. App. Vol. Ill at 520, 527-29. Ultimately, he concluded that “there was insufficient objective medical evidence to support any *698of the exclusionary diagnoses ... including current objective evidence of radiculopathy.” Id. at 528. Referring specifically to the March 9, 2006 EMG report, he noted that “[o]ld diagnostic reports indicating evidence of ‘chronic denervation’ at L5 do not support current or continuing evidence of radiculopathy.” Id.
That Ms. McClenahan’s physician, Dr. Drye, drew a contrary conclusion doesn’t suffice to render MetLife’s decision arbitrary and capricious. After all, MetLife has no obligation to “accord special weight to the opinions of a claimant’s physician; nor may courts impose on plan administrators a discrete burden of explanation when they credit reliable evidence that conflicts with a treating physician’s evaluation.” Black & Decker Disability Plan v. Nord, 538 U.S. 822, 834, 123 S.Ct. 1965, 155 L.Ed.2d 1034 (2003). Of course, it’s equally true that insurers “may not arbitrarily refuse to credit a claimant’s reliable evidence, including the opinions of a treating physician.” Id. But in this case, MetLife didn’t disregard Dr. Drye’s opinion. To the contrary, it sought out his expertise in light of inconsistent medical conclusions from other physicians; analyzed and relied on the conclusions of various physicians (both in-house and independent); requested the claimant to supplement her medical file throughout the administrative appeals process; and considered opinions from the other side. See, e.g., Holcomb, 578 F.3d at 1193 (denial of benefits was not an abuse of discretion, even though the insurer “received a large volume of reports, letters, imaging studies, and exams that were not entirely consistent”). MetLife’s ultimate decision may not have been the one Ms. McClenahan sought, and it may not have been the only decision available to the company on the record created by the parties. But neither can we say that its decision was an unreasonable one unsupported by substantial evidence. And more than that our precedent does not require.
In reply, Ms. McClenahan places heavy emphasis on Dr. Stephens’s March 9 report, arguing that MetLife gave it unduly short shrift. But to establish Ms. McClenahan’s entitlement to benefits after March 13, MetLife’s policy required evidence establishing her continuing disability on or after that date. Dr. Stephens’s examination establishing that she suffered from radiculopathy at an earlier date, even a reasonably proximate earlier date, may very well suggest she continued to suffer from the disorder on or after March 13, but it cannot suffice to compel such a conclusion. After all, patients sometimes do improve. Besides, even considering the March 9 report on its own terms, it rated Ms. McClenahan’s overall motor functions as normal and suggested that denervation near her lower spinal column was “likely,” not certain. Aplt.App. Vol. Ill at 564. Given the equivocal nature of this finding and its timing, we cannot disagree with the district court’s conclusion that it was reasonable for MetLife to rely more heavily on other physicians who found no evidence of radiculopathies present on or after March 13.
To be sure, Ms. McClenahan also argues that Dr. Monkofsky’s statement — that there was no “current or continuing evidence of radiculopathy” — suggests he only looked for radiculopathies as of October 2007, the time he examined her medical records. And this is a problem, Ms. McClenahan argues, because MetLife itself indicated that it was her condition as of March 13, 2006, not some later date, that mattered for determining her entitlement to continued disability benefits. But reading the physician’s report as a whole, we can’t draw the inference Ms. McClenahan suggests. Cf. Holcomb, 578 F.3d at 1194 (“Because [claimant’s] argument *699overstates the significance of one sentence in a large administrative record, it is unavailing.”). In responding to MetLife’s inquiry, Dr. Monkofsky’s report expressly took account of Ms. McClenahan’s medical data available from October 2003 through October 2007 — belying any contention that he was only concerned about the presence of radiculopathies in October 2007.
C
Finally, Ms. McClenahan argues that the district court erred by denying her request to supplement the administrative record with a medical report dated March 19, 2008. Again, we cannot agree. Before closing her file in November 2007, MetLife provided Ms. McClenahan with plenty of notice and time to submit evidence of radiculopathies “beyond” March 13, 2006, but the evidence she submitted only suggested prior radiculopathy. Aplt. App. Vol. IV at 607 (“April 6, 2007 letter”) (asking her to “include copies of all medical documentation to support that [she] has remained disabled from any occupation beyond March 13, 2006 due to” radiculopathies). In these circumstances, the district court properly ruled that it was obligated to reject Ms. McClenahan’s effort to introduce new evidence in the trial court proceedings and to limit its review to the administrative record. Indeed, our binding precedent offered the district court no other possible path. See Hall v. UNUM Life Ins. Co. of Am., 300 F.3d 1197, 1200-01 (10th Cir.2002); Metzger v. UNUM Life Ins. Co. of Am., 476 F.3d 1161, 1166 (10th Cir.2007) (“Permitting a claimant to receive and rebut medical opinion reports generated in the course of an administrative appeal ... would set up an unnecessary cycle of submission, review, re-submission, and re-review.”).2
The judgment of the district court is affirmed.