McKAY, Circuit Judge.

In this action brought under the Price-Anderson Act, 42 U.S.C. § 2210, we must decide whether Plaintiffs alleging they suffered cancer due to exposure to radiation from Defendants' uranium mill have made a sufficient showing of causation under New Mexico law to survive summary judgment. We first determine the test for causation in this context, then evaluate whether the evidence submitted by Plaintiffs was sufficient to satisfy this test for summary judgment purposes.

This action was originally brought by several plaintiffs who alleged they or the decedents they represented suffered from a large variety of injuries and diseases as a result of exposure to radioactive and non-radioactive hazardous substances released from Defendants' uranium milling facility in Cibola County, New Mexico. The district court entered a scheduling order requiring each plaintiff to produce expert affidavits making a prima facie showing of harmful exposure and specific causation for each alleged injury, but only three plaintiffs—the appellants in this action— did so. The district court dismissed the other twenty-five plaintiffs from the action with prejudice, and that dismissal is not contested in this appeal. As for the three Plaintiffs involved in this appeal, their experts opined these Plaintiffs' exposure to radiation from Defendants' operations was a substantial factor contributing to each of them developing cancer. The district court concluded that New Mexico law required a showing of but-for causation and that Plaintiffs' expert affidavits failed to meet that showing. The court therefore granted summary judgment to Defendants on Plaintiffs' claims. This appeal followed.

We review the district court's grant of summary judgment de novo, applying the same legal standard applicable to the district court. See Padhiar v. State Farm Mut. Auto. Ins. Co., 479 F.3d 727, 732 (10th Cir.2007). Under this standard, summary judgment is only warranted "if the pleadings, the discovery and disclosure materials on file, and any affidavits show that there is no genuine issue as to any material fact and that the movant is entitled to judgment as a matter of law." Fed. R. Civ. Pro. 56(c)(2). Because the alleged "nuclear incident" at issue in this case occurred in New Mexico, Plaintiffs' claims are governed by the substantive law of New Mexico. See 42 U.S.C. § 2014(hh). The parties have not cited, and we have not found, any New Mexico authority expressly considering the question of causation in a toxic torts case. We therefore look to New Mexico law on causation generally to attempt to predict how the New Mexico Supreme Court would rule if faced with this issue. See Wankier v. Crown Equip. Corp., 353 F.3d 862, 866 (10th Cir. 2003).

In New Mexico, as is universally the case, a tort plaintiff must demonstrate the defendant's actions caused the plaintiff's injury. Since 1892, New Mexico has generally required the plaintiff to make a showing of "that cause which, in natural and continued sequence, unbroken by any efficient, intervening cause, produced the result complained of, and without which that result would not have occurred." Lutz v. Atl. & Pac. R.R. Co., 6 N.M. 496, 30 P. 912, 916 (1892) (internal quotation marks omitted) (emphasis added). Stated differently, New Mexico courts have indicated that a tort plaintiff must demonstrate there is "a chain of causation initiated by some negligent act or omission of the defendant, which in legal terms is the cause in fact or the `but for' cause of plaintiff's injury." Chamberland v. Roswell Osteopathic Clinic, Inc., 130 N.M. 532, 27 P.3d 1019, 1023 (N.M.Ct.App.2001). Plaintiffs argue, however, that this general rule is not applicable in cases such as toxic tort cases that involve multiple potential contributing causes. They argue that causation in such cases may be proven through a substantial factor test, without regard to whether the injuries would likely have occurred in the absence of the defendant's actions.

Plaintiffs first contend New Mexico's uniform jury instructions make clear that but-for causation is not required in every case. New Mexico Civil Uniform Jury Instruction 13-305, in its description of causation, places brackets around the phrase "and if injury would not have occurred without it." These brackets mean this phrase might "need to be adapted in accordance with the acts in each particular case." 4 New Mexico Rules Annotated, 5 (2009). The commentary to the jury instructions, however, lists only two situations in which the but-for clause may be unnecessary or inappropriate: (1) cases involving alternative liability, such as the classic case of Summers v. Tice, 33 Cal.2d 80, 199 P.2d 1 (1948), and (2) cases in which "multiple acts each may be a cause of indivisible injury regardless of the other(s)," as described in § 432(2) of the Restatement (Second) of Torts. N.M. Unif. Jury Instruction Civ. 13-305, comm. cmt. (2009). Neither of these examples is applicable in the instant case.

The first exception to but-for causation described in the commentary to the jury instructions applies to "the unusual circumstances" in which two or more defendants engage in simultaneous or nearly identical negligent acts but only one of these acts causes the injury complained of, thus making it difficult or impossible for the plaintiff to prove which defendant caused the harm. Roderick v. Lake, 108 N.M. 696, 778 P.2d 443, 447-48 (1989), overruled on other grounds by Heath v. La Mariana Apts., 143 N.M. 657, 180 P.3d 664, 670 (2008).1 Plaintiffs do not argue that the alternative liability theory applies to this case, nor do we see a basis for alternative liability where only one potential wrongdoer has been identified and the injury may simply have resulted from natural causes. Cf. Menne v. Celotex Corp., 861 F.2d 1453, 1466 (10th Cir.1988) (describing a limited exception to the general requirement that all potential defendants must be before the court in an alternative liability situation, where it is clear the plaintiff's injury was caused by asbestos and all or substantially all available and identifiable implicated asbestos manufacturers are before the court).

The second exception described in the commentary to the jury instructions specifically refers to Section 432(2) of the Restatement (Second) of Torts. This exception involves situations in which multiple sufficient causes result in an indivisible injury—for instance, when two independently-set forest fires converge to burn a building, where either fire alone would have caused the same harm. See Restatement (Second) of Torts § 432 cmt. d, illus. 3 (1965); Restatement (Third) of Torts, § 27 cmt. a, illus. 1 (2005). In such circumstances, each act may be considered a cause of the plaintiff's injuries because it would, either alone or as a necessary part of a combination of other factors, have caused the harm in the absence of the coincident act. We recently considered this exception in June v. Union Carbide Corp., 577 F.3d 1234 (10th Cir.2009) (applying Colorado law), and interpreted Section 432(2) to require the plaintiff "to show that the conduct (or the causal set of which it is a necessary part) would in fact have caused the injury" if the competing cause(s) had not been present. Id. at 1243. Specifically, where a plaintiff alleges the defendant's actions caused the plaintiff to suffer from cancer, we held that the Restatement exception to but-for causation provides that "[o]nly a substance that would have actually (that is, probably) caused the cancer can be a factual cause without being a but-for cause," and that a defendant cannot be held liable simply on the basis that its product could have potentially caused or contributed to that specific type of cancer. Id. While New Mexico courts could of course interpret this exception differently, we see no basis in current New Mexico law for extending this exception beyond the limited bounds we described in interpreting the Restatement view under Colorado law in June.

Plaintiffs also argue that several New Mexico cases indicate a substantial factor test, rather than the but-for test, is appropriate in circumstances where there are multiple potential contributing causes for an injury. However, the cases cited by Plaintiffs, with one possible exception, are fully consonant with the Restatement view of causation as we explained it in June. In our view, New Mexico cases using the phrase "substantial factor," like the Colorado cases we discussed in June, do not create an alternative ground to but-for causation. See, e.g., Collins v. Perrine, 108 N.M. 714, 778 P.2d 912, 916 (N.M.Ct. App.1989) (using the phrase "substantial factor" in explaining and adopting the Restatement view of independent intervening cause); see also State v. Montoya, 133 N.M. 84, 61 P.3d 793, 799 (2002) (noting in the criminal law context that "it is difficult to comprehend how a person's conduct can ever be a substantial factor in causing a result that was going to occur when it did even without his contribution" (internal quotation marks omitted)).

As for the one possible exception, Tafoya v. Seay Bros. Corp., 119 N.M. 350, 890 P.2d 803 (1995), the plaintiff in that case argued, in essence, that the defendant's negligence was a but-for cause of the extent of the plaintiff's injury, even if it was not a but-for cause of the injury itself. Specifically, the plaintiff in Tafoya alleged a large rock popped out from under the tires of a dump truck and hit him "so hard that it propelled him forward ten to fifteen feet from where he had been standing and knocked him to the ground," causing him to suffer from a herniated cervical disc and a torn shoulder rotator cuff. Id. at 804. Although the truck had been negligently driven at an excessive speed for the road conditions, the trial court dismissed the case because there was no evidence that driving the truck slowly would have prevented the rock from striking the plaintiff. Id. at 804-05. However, the New Mexico Supreme Court stated the plaintiff was only required to "present some evidence, even circumstantial evidence, that the excessive speed contributed to [his] injuries." Id. at 805. The court then held that the plaintiff established a prima facie case of proximate cause by presenting evidence of the extent of his injuries, "the common knowledge that increased speed results in increased force," and evidence the rock came from under the tires of a truck which was traveling at a dangerous speed. Id. at 806. As we read this holding, the important point in Tafoya was that the plaintiff presented evidence he would not have been injured to the same extent were it not for the defendant's negligence, even if he might have suffered some injury without this negligence. Thus, even if the defendant's negligence was not the but-for cause of the incident itself, it was the but-for cause of the aggravation of the injury suffered by the plaintiff. We do not read this case as supplanting New Mexico's general but-for causation test with a contributing factor test, particularly in cases where there is no allegation that the injury was compounded by the tortfeasor's actions.

Finally, Plaintiffs argue that the requirement to show but-for causation will cut off virtually all relief for toxic tort plaintiffs because an honest scientific expert will never be able to state as a concrete fact that a plaintiff's injury would not have occurred were it not for exposure to the defendant's product. We note, however, that but-for causation in this context, as in other contexts, does not require proof to an absolute certainty. Although a plaintiff must prove but-for causation, the burden of proof in medical tort cases requires only "proof to a reasonable degree of medical probability," Alberts v. Schultz, 126 N.M. 807, 975 P.2d 1279, 1286 (1999), which "connote[s] proof that a causal connection is more probable than not," id. at 1287. Thus, as we interpret New Mexico law, a toxic tort plaintiff must demonstrate only to a reasonable degree of medical probability—not as a certainty—that exposure to a substance was a but-for cause of the injury or would have been a but-for cause in the absence of another sufficient cause. While we ultimately conclude Plaintiffs did not make this showing in the instant case, we are not persuaded this requirement is so insurmountable that New Mexico would create a toxic-tort exception to its general rule of but-for causation. Of course, New Mexico may certainly create such an exception if it sees fit, but we are not persuaded it would be appropriate for us to read such an exception into New Mexico law in this case. Cf. June, 577 F.3d at 1245 (concluding "it would be too adventurous on our part to assume that Colorado would depart from the Restatements" and deviate from the but-for causation standard).

We see no basis in New Mexico law for creating an exception to but-for causation simply because a case involves toxic torts, nor have Plaintiffs established any other basis for an exception to but-for causation in this case. We accordingly turn to the question of whether Plaintiffs made a sufficient showing of but-for causation to survive summary judgment.

Plaintiffs argue that causation is established through the following evidence: (1) Dr. Owen Hoffman's calculation of each Plaintiffs' "assigned share," which reflects "the proportion of individuals in a heterogeneous population who are alike in all aspects of age, gender, exposure history, and dose, who would not have been diagnosed with their cancer had they not been exposed" and is "an indirect statistical indicator of cancer causation for a given individual exposed to radiation" (Appellants' App. at 111); (2) Dr. James Ruttenber's testimony, based on the assigned share values calculated by Dr. Hoffman, that "a medical expert could consider radiation exposures from the Homestake mill to be a substantial contribution to the causes of the liver cancer, thyroid cancer, and bladder cancer of Plaintiffs Harriet Wilcox, Darlene Cowart-Serna, and Marcella Sally Hartman, respectively" (Appellants' App. at 54-55); (3) Dr. Inder Chopra's opinion that Ms. Serna was "exposed to potentially carcinogenic doses of radionuclides and her exposure was of sufficient magnitude, duration and intensity that it may have been a substantial contributing factor contributing to the onset, growth and/or progression of her thyroid cancer" (Appellants' App. at 45) and his supplemental language that Ms. Serna's exposure "was a contributing factor to her cancer" (Appellants' App. at 114); and (4) Dr. Robert Gale's opinion that "to a reasonable medical probability exposure to ionizing radiations as a consequence of Defendants' operations was a substantial factor contributing to each plaintiff developing cancer" (Appellants' App. at 50). However, as we held when we considered virtually identical statements made by the same experts in June, this evidence is simply insufficient to meet Plaintiffs' burden of making a prima facie case that Defendants' operations either (1) were a but-for cause of their cancer, either alone or as a necessary part of a combination of different factors, or (2) would have been such a but-for cause were it not for another sufficient coincident cause.2 See June, 577 F.3d at 1246. We also note Plaintiffs have neither alleged nor presented evidence that exposure to Defendants' radiation resulted, more probably than not, in the aggravation of their cancer symptoms. Therefore, to the extent Tafoya alters the but-for test in situations where a defendant's actions aggravate but do not cause an injury, it is not applicable in this case. For substantially the same reasons we explained in June, we conclude the district court correctly granted summary judgment to Defendants based on Plaintiffs' failure to make a sufficient showing of but-for causation. We therefore AFFIRM the district court's entry of summary judgment in favor of Defendants.

LUCERO, J., concurring in part.

I agree that New Mexico law requires plaintiffs to demonstrate but-for causation, and that they did not create a genuine factual dispute on that issue. I accordingly join the majority's discussion of the appropriate legal standard and concur in its judgment. I write separately to elaborate upon the manner by which plaintiffs in toxic torts cases can satisfy the but-for causation standard and to state my disagreement with the majority's characterization of June v. Union Carbide Corp., 577 F.3d 1234 (10th Cir.2009).

I

Although I agree generally with the majority's recitation of the facts, additional context is helpful in evaluating statements made by three of the plaintiffs' experts. First, Dr. F. Owen Hoffman calculated an "assigned share" figure for plaintiffs Wilcox, Hartman, and Serna. He defined assigned share as "an empirical expression of the proportion of individuals in a heterogenous population who are alike in all aspects of age, gender, exposure history, and dose, who would not have been diagnosed with their cancer had they not been exposed." Dr. Hoffman's numbers were based on the dose of radiation each person's diseased organ received from the defendants' mine. At a 95% confidence interval, the assigned shares ranged from 0.5% to 41% for Hartman, from 0.073% to 25% for Serna, and from 3.3% to 80% for Wilcox. In other words, Dr. Hoffman could state with a relatively high degree of statistical probability that the chance Hartman's cancer had been caused by radiation exposure from the mine was somewhere between 0.5% and 41%, and so on for Serna and Wilcox. The mean assigned shares were 12% for Hartman, 4.9% for Serna, and 45% for Wilcox.

Second, Dr. Robert Peter Gale opined on the cause of Wilcox's and Hartman's cancers as follows:

(emphasis omitted).

Finally, Dr. Inder J. Chopra separately considered the cause of Serna's cancer. After examining Serna's medical records, family and social histories, details of her exposure to radiation from the mine, and "other background risk factors" for cancer, Dr. Chopra opined, "Serna's exposure to radioactivity was of sufficient magnitude, duration and intensity that, in my expert opinion, rendered it a substantial factor to her development of thyroid cancer." He further averred that he "discovered no other overwhelming risk factors that may have caused Ms. Serna's [cancer]."

II

A

The majority opinion treats the panel decision in June as controlling our application of the but-for causation standard. (Majority Op. 1170.) I disagree. The June majority ruled in favor of the defendants with respect to but-for causation on procedural, not substantive, grounds:

577 F.3d at 1246-47. In dissent, Judge Holloway underscored that the June majority's decision was based on waiver: "I acknowledge that . . . plaintiffs did not timely argue before the district judge that they had produced sufficient evidence of but-for causation. . . . However, . . . I am convinced that special circumstances [warranting consideration of a waived issue] are present here." Id. at 1253 (Holloway, J., concurring in part, dissenting in part).

Unlike those in June, these plaintiffs have consistently argued that their proffered expert evidence satisfies New Mexico's but-for causation standard. The plaintiffs stated below that "[i]t is not necessary (nor is it possible) to opine that but for exposures [plaintiffs] would never have suffered an injury in the same category." Contrary to the district court's suggestion, this statement does not amount to a "free[] acknowledge[ment] that the cause of cancer is unknown." Instead, it was a reminder to the district court that medical experts speak in probabilities, not certainties. Because plaintiffs preserved the but— for argument below, the June majority should not control our analysis.

B

We lack clear guidance from New Mexico courts with respect to the evidence required to meet the but-for causation standard in toxic tort cases. Tenth Circuit precedent similarly sheds little light on the issue. Faced with this lacuna, I would apply the principles fleshed out in Judge Holloway's well-reasoned June partial dissent in evaluating such expert evidence.1

In June, a medical expert (coincidentally, the same Dr. Chopra who provided sworn statements in this case) testified that the defendants' uranium facility was "a substantial factor contributing" to the plaintiffs' illnesses. Id. at 1254. But in June, Dr. Chopra went much further, explaining:

Id. Moreover, he specifically opined that "[i]t is more probable than not that [the defendants' uranium facility] of itself was a substantial contributing factor to each Plaintiff's thyroid disease, without which the Plaintiff's respective thyroid disease would not have happened." Id. (emphases and first alteration in original).

In Judge Holloway's view, which I share, Dr. Chopra's use of the phrase "substantial factor" was immaterial. He could have used any term whatsoever to speak about causation; the incantation of magic words is not required, provided that an expert's opinion roughly corresponds to the legal concept of but-for causation. Dr. Chopra's testimony in June is a good example of the type of evidence that clears this hurdle. In addition, a plaintiff can prevail by adducing evidence suggesting that a defendant's polluting activities worsened or hastened the plaintiff's illness. Cf. Tafoya v. Seay Bros. Corp., 119 N.M. 350, 890 P.2d 803, 806 (1995) (evidence that a truck was travelling at a dangerous speed, propelling a rock into plaintiff with "increased force" sufficient to establish but— for causation).

But in this case, plaintiffs did not produce such evidence. Drs. Gale and Chopra did not define what they meant by the term "substantial factor." As the expert testimony in June makes quite clear, "substantial factor" is not a term of art from which sine qua non causation can be inferred; instead, it takes on a meaning that can differ depending on the "particular case." 577 F.3d at 1254.

When their opinions are read as a whole, neither Dr. Gale nor Dr. Chopra concludes that plaintiffs' cancers would not have arisen but for their exposure to defendants' radiation. Dr. Gale merely opined that defendants' mining operations were a substantial factor leading to Wilcox's and Hartman's cancers, "[a]ssuming" that no other factor "could, as a sole cause over-whelmingly explain each plaintiff having the cancer specified." Similarly, Dr. Chopra opined that defendants' mining operations were a substantial factor in the development of Serna's cancer, and that he knew of "no other overwhelming risk factors" that could cancel out the uranium mine's role. Neither of these qualified statements could reasonably be read to rule out other potential causes of the plaintiffs' cancers. Rather, the experts' qualifiers here merely highlight that no other known factor accounted for the uranium mine's role in cancer formation—whatever the uranium mine's role actually was.

Of course, an expert need not opine that the defendants' actions were the sole cause of the injury. As a general rule, a plaintiff demonstrates but-for causation by showing that his injury would not have been sustained absent the defendant's negligence, and there can be multiple but-for causes of a plaintiff's injury. See Restatement (Second) of Torts § 432. But the statements submitted to the district court do not claim that plaintiffs' cancers probably would not have occurred but for defendants' actions. Further, the experts' reference to Dr. Hoffman's mean assigned share values— all below 50%—suggests that they did not intend the term "substantial factor" to be the functional equivalent of but-for causation. Even Dr. Chopra's supplemental affidavit, submitted after the causation issue was squarely raised, does not support such an inference.

III

In toxic tort cases, the causation issue presents a difficult hurdle for plaintiffs and their experts. But that hurdle is not insurmountable. As Judge Holloway's considered partial dissent in June demonstrates, an expert need not revert to legalisms or rule out all other potential causes of a plaintiff's injury. Instead, the totality of an expert's testimony must convey that he or she believes the plaintiff probably would not have gotten sick—or the illness would have been later-developing or less severe—but for the defendant's negligence. Unfortunately for Wilcox, Hartman, and Serna, their experts conveyed no such belief. Therefore, I join the majority in affirming the district court's grant of summary judgment to the defendants.

HOLMES, Circuit Judge, concurring.

I write separately because my reading of the record reveals that Plaintiffs have waived any argument that the but-for standard of causation does not apply. Although I have no reason to question the majority's view that the New Mexico Supreme Court would apply the but-for standard rather than the substantial-factor test, I would not reach Plaintiffs' argument on this point and offer no definitive opinion on the matter here. Instead, I would hold that the but-for standard is the applicable causation standard in this case—Plaintiffs having waived any argument to the contrary—and would proceed directly to the question of whether Plaintiffs offered sufficient evidence under that standard to survive summary judgment. On that question, I agree fully with the majority's reasoning and its ultimate conclusion that Plaintiffs did not offer sufficient evidence.

On appeal, Defendants contend that Plaintiffs did not argue before the district court that the substantial-factor test applied. I agree. In opposing Defendants' motion for summary judgment, Plaintiffs did not ask the district court to apply the causation standard that they now advance on appeal—viz., the substantial-factor test. Rather, Plaintiffs contended that the but— for standard applied and urged the district court to hold that their evidence satisfied that standard. For example, they stated that "New Mexico requires both a showing that an event is a contributing factor and a cause without which an injury would not have occurred, or more commonly, 'but for' causation."1 Aplts. App. at 77 (emphasis added). Plaintiffs argued that, even if their experts did not use "the magic words, `but-for,'" id. at 78, "each opinion conceptually and legally satisfies `but for' causation," id. at 79. Most notably, Plaintiffs explicitly acknowledged that the significant-factor test (which apparently Plaintiffs viewed as synonymous with the substantial-factor test) was not the prevailing test in New Mexico. See id. at 86 (stating that "[w]hile it is true that the `significant factor test' is a legal test accepted in multiple jurisdictions, simply because Dr. Gale used that terminology more familiar in other jurisdictions, it does not somehow discredit his causative opinion in New Mexico" (emphasis added)). Therefore, in granting summary judgment in favor of Defendants, the district court correctly read Plaintiffs' memorandum as advancing the but-for standard.2 See Aplts. App. at 139 ("Their expert affidavits satisfy the but-for test for causation, according to Plaintiffs." (emphasis added)).

Because Plaintiffs did not argue before the district court that the substantial-factor test governed the determination of whether Plaintiffs' cancer was caused by radiation that Defendants emitted, we should not consider Plaintiffs' attempt for the first time on appeal to convince us to apply that test. See Tele-Commc'ns, Inc. v. Comm'r, 104 F.3d 1229, 1232-33 (10th Cir.1997); Lyons v. Jefferson Bank & Trust, 994 F.2d 716, 720-21 (10th Cir. 1993). Plaintiffs have waived that argument and I can discern no "most unusual circumstances" that would warrant an exercise of our discretion to decide this question despite Plaintiffs' waiver. Lyons, 994 F.2d at 721; accord Rademacher v. Colo. Ass'n of Soil Conservation Dists., Med. Benefit Plan, 11 F.3d 1567, 1572 (10th Cir.1993); see also Hicks v. Gates Rubber Co., 928 F.2d 966, 970-71 (10th Cir.1991) (explaining the policy rationale for declining to consider arguments raised for the first time on appeal).

Because I would hold Plaintiffs to their waiver, I would proceed directly to the second question on appeal—whether Plaintiffs experts' affidavits made a sufficient showing of but-for causation to survive summary judgment. I agree fully with the majority's conclusion that they did not do so.