For nearly 40 years, Harold Huffman worked for the Union Pacific Railroad. He claims that injuries to his knee, diagnosed after his retirement, were partly the result of the railroad’s negligence. A jury found such a connection and awarded damages. We conclude that no evidence was introduced to connect the worker’s specific condition to the work that he performed. We therefore REVERSE and REMAND for entry of judgment.
FACTS
This suit was filed under the Federal Employers Liability Act (FELA, or “the *415Act”). 45 U.S.C. § 51 et seq. Huffman filed the suit on February 21, 2007. The claims at trial were that the Union Pacific was negligent in failing to: (1) provide a reasonably safe work environment,- (2) give adequate warnings and training to its employees regarding the risks and harms associated with their work, (3) perform ergonomic screening or appropriate job analysis of the jobs performed by Huffman, and (4) implement a comprehensive plan to reduce the avoidable physical stresses in employees’ job duties.
Huffman was first employed in 1965 and retired in 2004. He claims that the repetitive physical demands of his work resulted in the cumulative trauma injury1 of knee osteoarthritis. Evidence supported that a frequent walker along a railroad roadbed, who must occasionally leap off moving trains, and who otherwise does what those who work on the railroad have long done, is subject to musculoskeletal disorders. The principal issue on appeal is whether jurors could infer that Huffman’s osteoarthritis, an injury that can be caused by everyday activities, was caused by the railroad’s failure to train him on how to perform his tasks in an ergonomically optimal way.
Huffman started his career as a brakeman. From 1970 until his retirement, he worked as a conductor. That position placed him in charge of the train and the work of others. Huffman testified that even as a conductor, as late as 1993, he still performed job functions of a trainman.
In 1965, there were separate positions for an engineer, conductor, fireman, head brakeman, rear brakeman, and switchman. Now, the brakeman, switchman, and conductor functions are performed by a “trainman.” The Union Pacific’s description of the physical requirements of the trainman position is this:
Must be able to exert musculoskeletal strength — or muscular strength sufficient to push, pull, and lift/earry weights of up to 25 pounds frequently, and 50 pounds occasionally, and to assist in the infrequent movement of weights up to 83 pounds. Strength, coordination, and balance sufficient to step on and off moving equipment at speeds of three miles per hour or less. Walk frequently on ballast across uneven terrain for short distances and infrequently for distances up to one mile. Remain seated or standing for over a half work shift with the opportunity to change position for comfort. Climb stairs and ladders occasionally. Perform activities that infrequently require bending or stooping. Hold on to a ladder while occasionally riding a slowly moving train.
Often the only employees working on a train were the engineer and trainman/conductor. Huffman jumped off trains going faster than three miles per hour on a few occasions. He also walked on the loose and uneven ballast (the rocks helping secure the tracks) for over a mile at a time, causing him to slip and slide. The mainline ballast, ie., the rocks on the miles of mainline tracks, were larger than the ballast in railroad yards. Huffman frequently had to lift and move more than 50 pounds of equipment, bend over to grab switch levers, and lift switches.
There was evidence that the Union Pacific developed a training program to reduce injuries resulting from the cumulative trauma common to railroad workers. Huffman testified that he never received the benefits of this program. He was not *416trained on the proper way to walk on ballast or on the safe method of getting on or off moving equipment. At trial, several video clips developed by another railroad were played for jurors that showed the proper way for trainmen to minimize stress to muscles, bones, discs, and joints.2 Huffman’s counsel argued that these videos demonstrated that the Union Pacific knew the ailments from which Huffman suffered could result from improper work habits. Huffman testified that he had never seen those or similar videos. None of those videos or other evidence stated that osteoarthritis could result.
After Huffman’s retirement, he was diagnosed with osteoarthritis in his knees. This was the only continuing injury for which evidence was presented to the jury and therefore the only injury at issue. At the time of trial, he was not taking prescription medicine and had neither received nor scheduled knee surgery. His knee pain had limited his ability to hunt, dance, walk, and care for his lawn. He enjoys hunting and fishing, which he can still do, but his endurance during these activities is diminished due to his knee pain. His hunting has been curtailed because of the difficulty of extensive walking.
At trial, the Union Pacific emphasized other explanations for Huffman’s knee problems. He was 65-years old, weighed about 300 pounds, and was 5'9" tall. There was no evidence that his weight was different before his retirement. There was evidence that someone Huffman’s size was 15 times more likely to develop his particular knee problems than someone who is not overweight.
After Huffman rested his case, the Union Pacific moved for a directed verdict on the grounds that Huffman had failed to present competent evidence on causation linking the Union Pacific’s alleged negligent conduct to Huffman’s knee injury. In arguing that there was sufficient evidence even without expert testimony, Huffman cited Gutierrez v. Excel Corp., 106 F.3d 683 (5th Cir.1997). We will discuss the precedent below. The district court denied the motion.
The jury returned a verdict in Huffman’s favor upon finding that the Union Pacific’s negligence was one of the causes of Huffman’s osteoarthritis. Huffman was awarded $606,000 in damages. The Union Pacific moved for a judgment as a matter of law, arguing there was insufficient evidence on causation. Once again, Huffman argued that plaintiffs bringing negligence claims for cumulative trauma disorders were not required to “present medical or expert testimony specifically stating that there is a direct causal link between a defendant’s actions and a plaintiffs injury.” The Union Pacific also moved for a new trial, arguing the damages were excessive or for a remittitur. The district court denied both motions. This appeal timely followed.
DISCUSSION
Under FELA, an injured railroad employee may recover damages for “injury or death resulting in whole or in part from the negligence” of the railroad. 45 U.S.C. § 51. FELA provides the exclusive remedy for a railroad employee engaged in interstate commerce whose injury resulted from the negligence of the railroad. Rivera v. Union Pac. R.R. Co., 378 F.3d 502, 507 (5th Cir.2004). Any trial necessary to resolve the claims is before a jury. Atl. & Gulf Stevedores, Inc. v. Ellerman Lines, Ltd., 369 U.S. 355, 360, 82 S.Ct. 780, 7 L.Ed.2d 798 (1962).
*417FELA eliminated a variety of traditional defenses, such as the fellow servant rule, the assumption of the risk defense, and the doctrine of contributory negligence. Consol. Rail Corp. v. Gottshall, 512 U.S. 532, 542-43, 114 S.Ct. 2396, 129 L.Ed.2d 427 (1994); 45 U.S.C. §§ 51, 53-55. A railroad is charged with providing a reasonably safe work environment for its employees. See Urie v. Thompson, 337 U.S. 163, 179 n. 16, 69 S.Ct. 1018, 93 L.Ed. 1282 (1949). Negligence within the meaning of FELA exists if the defendant railroad “knew, or by the exercise of due care should have known” that its conduct was “inadequate to protect [the plaintiff] and similarly situated employees.” Id. at 178, 69 S.Ct. 1018.
On appeal, no issue is made of the validity of the jury’s determination that the Union Pacific was negligent. The jury was given evidence that the railroad was negligent by, along with other possibilities, failing to educate Huffman on less physically harmful methods of performing his everyday duties. Therefore, the workplace was not as safe as it reasonably could have been.
What is questioned is whether there was legally sufficient evidence to support that the railroad’s negligence, as the jury instructions put it, “contributed in any way to any injury or damage suffered by the plaintiff.” The Union Pacific earlier questioned whether that is the correct standard. The question was answered after oral argument in this case by the Supreme Court. Under FELA, a “defendant railroad ‘caused or contributed to’ a railroad worker’s injury ‘if [the railroad’s] negligence played a part- — -no matter how small — in bringing about the injury.’ ” CSX Transp., Inc. v. McBride, — U.S. —, 131 S.Ct. 2630, 2644, 180 L.Ed.2d 637 (2011) (alteration in original). Our previously stated standard is similar: liability arises if a railroad’s “negligence played any part — however small — in the development of his condition.” Rivera, 378 F.3d at 510 (quotation marks and citation omitted).
The Union Pacific organizes its appellate arguments into three issues: (1) a judgment as a matter of law should have been granted because of the absence of evidence of causation; (2) the district court erred in its instruction to the jury as to the necessary degree of causation; and (3) the damage award was excessive. In light of our conclusions, we need not address the last issue.
Before discussing the first issue, we review Huffman’s argument that the railroad is judicially estopped from contesting the sufficiency of the evidence.
A. Judicial Estoppel
Huffman contends that the Union Pacific is judicially estopped from arguing the injuries are not work related. Estoppel allegedly arises because the railroad sought in the district court to have the suit dismissed on the basis that Huffman had known for years that his osteoarthritis was work related and had not brought suit until after the applicable statute of limitations had run. The Union Pacific was granted a jury instruction that the railroad would have no liability if Huffman knew for more than three years before bringing his claims that his knees were injured due to work-related causes.
Judicial estoppel protects the legal system, not the parties, and requires a party not take intentionally self-contradictory positions at different points in a case to obtain an unfair advantage. In re Coastal Plains, Inc., 179 F.3d 197, 205-06 (5th Cir.1999). Required is that a position taken at one time in a suit be clearly inconsistent from that taken at another, *418and the party’s earlier position must have been accepted by the court. Id. at 206.
The simple answer is that the Union Pacific convinced neither the district court nor the jury that its position was correct. The slightly more involved answer is that both in pleadings and in presentation of evidence, a party may make alternative submissions. See Fed.R.Civ.P. 8(d)(2). At trial, there may be a requirement to elect remedies or for other reasons choose one particular theory, but no such circumstance exists here. Cf. Guy James Constr. Co. v. Trinity Indus., Inc., 644 F.2d 525, 529-30 (5th Cir.1981). Further, the Union Pacific could be seen as arguing that Huffman took a factual position that constituted a judicial admission barring his claim regardless of whether the railroad agreed with the assertion. See Heritage Bank v. Redcom Labs., Inc., 250 F.3d 319, 329 (5th Cir.2001).
As a matter of appellate practice, we also determine the issue was waived because the railroad’s allegedly inconsistent positions at trial were not challenged at that time. Nunez v. Allstate Ins. Co., 604 F.3d 840, 846 (5th Cir.2010). Only in an egregious case will we accept a judicial estoppel argument first raised on appeal that applies to a party allegedly taking inconsistent positions in the district court. Beall v. United States, 467 F.3d 864, 870 (5th Cir.2006). Far from being egregious, this example does not even clearly qualify as an inconsistency.
B. Proof of Causation Under FELA
1. General Principles
The Union Pacific moved for a judgment as a matter of law because of the alleged absence of evidence of causation. The motion was denied. That motion should be granted only when there is a “complete absence of probative facts to support the conclusion reached by the jury.” Rivera, 378 F.3d at 505. This standard “recognizes that the FELA is protective of the plaintiffs right to a jury trial.” Wooden v. Mo. Pac. R.R. Co., 862 F.2d 560, 561 (5th Cir.1989). On appeal, the railroad argues that denial of the motion was error.
We begin by summarizing some well-known concepts in order to focus on the point at which the parties join issue. The cause of action is one for negligence, which requires proof of breach of a standard of care, causation, and damages. Gottshall, 512 U.S. at 540, 114 S.Ct. 2396. The statute did not create a workers’ compensation system. Id. at 543, 114 S.Ct. 2396. If an injury has multiple causes, it is sufficient if the railroad’s “ ‘negligence played a part — no matter how small — in bringing about the injury.’ ” McBride, 131 S.Ct. at 2644. The burden of proof for the claim is on the employee. Norfolk S. Ry. Co. v. Sorrell, 549 U.S. 158, 163, 127 S.Ct. 799, 166 L.Ed.2d 638 (2007). Contributory negligence can reduce recovery under FELA, see 45 U.S.C. § 53, but there has been no issue of that in this case. The standard of care for both the employer and employee under FELA is one of ordinary prudence under the circumstances. Gautreaux v. Scurlock Marine, Inc., 107 F.3d 331, 338-39 (5th Cir.1997) (en banc).3
This appeal does not require the panel to wrestle with any of the preceding concepts. We must determine whether competent evidence was admitted of this minimal connection between railroad conduct and worker injury. There was no expert testimony to support a link between Huffman’s performance of his work duties in *419less than ergonomically optimal ways — a result of the railroad’s negligence — and the specific knee problem he suffers, which is osteoarthritis. Though there was no specific argument made in the district court that an expert witness was necessary, the issue was fully joined regarding whether there was sufficient evidence. Whether what is missing is expert testimony or some other form of evidence is not central to a sufficiency objection. In a pretrial order, the district court granted the Union Pacific’s motion to exclude the testimony of Huffman’s treating physician as an expert on causation. There is no challenge to that decision on appeal. Huffman therefore had no expert testimony as to causation at trial.
At oral argument before this court, the railroad’s attorney categorized Huffman’s claim as one based on a long-term, progressive, non-acute injury, with multiple potential causes. That characterization usefully focuses the issue of the kind of injury for which jurors had to find causation.
As part of our analysis of whether there was an evidentiary deficit, we begin with the observation that we have not specifically addressed when expert testimony is required in FELA cases. We reject Huffman’s assertion that our opinion in Rivera said anything about whether expert testimony was needed. After describing each party’s expert testimony, we concluded that there was some probative factual evidence to support the verdict for the plaintiff. Rivera, 378 F.3d at 510. The necessity of expert testimony was not before us in Rivera because both sides had provided it. Id. Here, neither party introduced expert testimony on causation. Both parties had a witness who would have given an opinion, but for different reasons, neither witness was allowed to present it.
Expert testimony is not needed in many if not most cases. As former Chief Judge Charles Clark noted, jurors are generally entitled to draw their own inferences from the evidence. Fontenot v. Teledyne Movible Offshore, Inc., 714 F.2d 17, 20 (5th Cir.1983). We rely on juror common sense to reduce the risk of exorbitant liability in FELA actions. McBride, 131 S.Ct. at 2643-44. “Jurors are supposed to reach their conclusions on the basis of common sense, common understanding and fair beliefs, grounded on evidence consisting of direct statements by witnesses or proof of circumstances from which inferences can fairly be drawn.” Fontenot, 714 F.2d at 20 (quoting Schulz v. Penn. R.R. Co., 350 U.S. 523, 527, 76 S.Ct. 608, 100 L.Ed. 668 (1956)). This general rule gives way, though, when conclusions as to the evidence cannot be reached based on the everyday experiences of jurors, making expert testimony necessary to evaluate the issue. Caboni v. Gen. Motors Corp., 398 F.3d 357, 361 (5th Cir.2005).
One of our sister circuits has held that expert testimony in FELA cases is not required if the connection between the negligence and the injury is fairly self-evident, such as that a broken leg would result from being struck by a motor vehicle. Moody v. Maine Cent. R.R. Co., 823 F.2d 693, 695-96 (1st Cir.1987). The causes of Huffman’s osteoarthritis are not of that clarity. Our question, then, is whether the totality of evidence would allow jurors reasonably to infer, without expert testimony, that negligence of the railroad contributed at least in some small way to his specific injury. A component of that analysis is whether there is something about Huffman’s claim of osteoarthritis in the knees that mandates the use of expert testimony.
2. Need for Expert Opinion on Causation
Expert testimony may be introduced even when not strictly necessary if it “will *420help the trier of fact to understand the evidence or to determine a fact in issue.” Fed.R.Evid. 702. Because the Union Pacific did not argue at trial that an expert was required, we will not decide whether an expert is necessary to show causation for this kind of physical condition. Instead, the perspective for our analysis is this: could jurors make a finding as to causation based on the evidence that they had?
In deciding whether this evidence was sufficient, Huffman urges us to consider a non-FELA case, in which we discussed proof of causation for a cumulative trauma injury in a common-law negligence action. Gutierrez v. Excel Corp., 106 F.3d 683, 686 n. 3 (5th Cir.1997) (noting that because the employer did not subscribe to the Texas workers’ compensation system, the suit was brought as a negligence action). The basis of liability under FELA is not simply that an injury occur while an employee is at work but also that some negligence occurred. Gottshall, 512 U.S. at 543, 114 S.Ct. 2396. The Gutierrez court allowed a jury in a negligence action to infer causation for a cumulative trauma injury even without expert witnesses. Gutierrez, 106 F.3d at 687.
In Gutierrez, cumulative trauma injury was “characterized as “wear and tear’ on the tissue surrounding joints, ligaments, and tendons. [It] refers not to one specific injury, but to numerous disorders caused by the performance of repetitive work over a long period of time.” Id. at 686. The theory of negligence in that case was that an electric knife used by assembly line workers was known to cause cumulative trauma disorders (“CTDs”), and the defendants failed to adopt safety measures to reduce the trauma. Id.
The risk factors for CTDs were shown through guidelines published by the Occupational Safety and Health Administration (OSHA), a division of the United States Department of Labor. Id. The guidelines identified the risk factors to the upper extremities present in the meat packing industry. Id. OSHA had published guidelines offering suggestions on minimizing those risk factors. Id. There was evidence that such measures had not been implemented at this workplace. Id. at 687-88.
The Gutierrez court also examined whether the kind of injuries that the OSHA documents stated would arise from cumulative trauma matched the kind of injuries suffered by the plaintiffs:
Injuries that may be classified as CTDs include, but are not limited to, carpal tunnel syndrome in the wrist, rotator cuff tendinitis in the shoulder, and nerve compression. While CTDs are generally not caused by any one specific traumatic event, there are certain risk factors associated with cumulative trauma, including repetition, force, vibration, cold, and posture.
Id. at 686.
The Gutierrez court held that there was evidence “that the working situation at the Whizard table was rife with conditions known to cause, or at least to be associated with, [CTD].” Id. at 688. The plaintiff whose proven injuries were the kind recognized as CTDs was granted a new trial. Id. at 690. As to her, “medical records suggest that her injury was ‘probably caused’ from repetitive work.” Id. Two plaintiffs whose specific injuries were ones that were not categorized as CTDs were not entitled to recover. Id. at 689.
The court stated that it was not necessary for “all plaintiffs bringing negligence claims for [CTDs to] present medical or other expert testimony specifically stating that there is a direct causal link between a defendant’s actions and a plaintiffs inju*421ry.” Id. First, plaintiffs could show they were negligently exposed to the “risk factors present in the work-environment that are known to be associated with [CTD].” Id. Second, plaintiffs could show through medical evidence that they suffered from the “particular injuries collectively referred to as [CTDs] that are caused by a combination of these risk factors.” Id.
Not requiring expert testimony in Gutierrez was consistent with Chief Judge Clark’s useful summary of the capabilities of jurors: a jury can draw inferences “on the basis of common sense, common understanding and fair beliefs, grounded on evidence consisting of direct statements by witnesses or proof of circumstances from which inferences can fairly be drawn.” Fontenot, 714 F.2d at 20 (quoting Schulz, 350 U.S. at 527, 76 S.Ct. 608). In Gutierrez, what substituted for expert testimony was proof that certain kinds of injuries result from certain kinds of work — that drew from expert analysis made by OSHA — and proof that the plaintiff engaged in that kind of work and suffered that kind of injury.
Our sister circuits have dealt with related issues. In determining liability under FELA, one of the other circuits recognized that if general causation evidence is offered, “a jury question certainly can be created even without expert testimony going directly to the question of specific causation.” Hardyman v. Norfolk & W. Ry. Co., 243 F.3d 255, 268 (6th Cir.2001). In that case, the court noted that evidence from the treating physician of the plaintiffs injury, the kinds of activities that result in that injury, the risk factors for the development of that injury, and evidence that those risk factors existed in the plaintiffs job was sufficient to create a jury question. Id. at 269.
A recent Seventh Circuit decision took a different approach. Myers v. Ill. Cent. R.R., 629 F.3d 639 (7th Cir.2010). There, the long-time railroad worker had job duties and cumulative trauma injuries similar to Huffman’s. Id. at 640-41. The worker argued that he merely needed to have an expert establish that conditions of his employment could cause certain kinds of injuries, and have his physician testify that these were the injuries that he had developed. The district court granted summary judgment for the railroad because no expert testified that his injuries were caused by the railroad’s negligence. Id. at 641-42. The Seventh Circuit affirmed, stating that “the nature of the trauma injuries that [the plaintiff] accumulated required expert testimony establishing specific causation.” Id. at 645. The court noted the debate over how plaintiffs must establish causation under FELA, and acknowledged the Sixth Circuit’s decision in Hardyman and cited this court’s decision in Gutierrez. Id. at 642-43.
The Union Pacific presses upon us a ruling from still another circuit in which the claim was based on a chemical exposure. See Claar v. Burlington N. R.R. Co., 29 F.3d 499 (9th Cir.1994). In addition to evidence that certain chemicals the plaintiff was exposed to at work caused the kind of injuries he suffered, the court also required expert testimony to show that the chemicals in fact did cause the plaintiffs injuries. Id. at 504.
The Claar court noted two Supreme Court decisions that “arguably stand for the proposition that, in order to create an issue for the jury, a FELA plaintiff need show only the mere possibility that a defendant’s actions caused his injuries.” Id. at 503-04 (citing Gallick v. Baltimore & Ohio R.R. Co., 372 U.S. 108, 83 S.Ct. 659, 9 L.Ed.2d 618 (1963); Lavender v. Kurn, 327 U.S. 645, 66 S.Ct. 740, 90 L.Ed. 916 (1946)). One of the cited precedents involved a railroad worker who became seri*422ously ill after an insect bite. Gallick, 372 U.S. at 109-10, 83 S.Ct. 659. Evidence of the existence of a stagnant pond next to where the employee worked, which provided a breeding place for insects, was sufficient circumstantial evidence to create a jury question on causation. Id. at 114, 83 S.Ct. 659. In the other case, there was little evidence upon which jurors could rely to decide between the deceased having been killed along the tracks by a metal hook hanging from a negligently maintained rail car or having instead been murdered. Lavender, 327 U.S. at 648-51, 66 S.Ct. 740. The jury was nonetheless free to make the inference that supported FELA liability. Id. at 652, 66 S.Ct. 740. In both cases, there was evidence that the death could have been caused by the condition created by the railroad. The jurors were free to choose among the possibilities.4
The argument Huffman advances is that Gutierrez-style proof of causation should be applied to his FELA claim, and if it is, he proved enough. As a common law negligence case, Gutierrez required a plaintiff to prove that an employer’s negligence was a substantial factor in bringing about the employee’s injury. This level of causation is much higher than in FELA cases. We certainly do not hold Huffman to this higher standard.
We do not see Gutierrez as an evidentiary model that we must decide may or may not apply in FELA cases. Instead, our task is simply to decide whether there was sufficient evidence for a jury to infer— based on their common sense and common understanding — that Huffman’s negligently-directed work with the Union Pacific contributed in any way to his development of osteoarthritis. The evidence need not show a substantial contribution but it must allow an inference that some contribution occurred. We now examine the evidence.
C. Evidence on Causation
Both parties had an expert available to testify regarding causation, but the experts were not allowed to testify for different reasons.
Huffman offered the testimony of Dr. Alan Smith, his treating physician. In a pretrial order, though, the district court granted the Union Pacific’s motion to exclude his testimony as an expert on causation. The district court held that Huffman’s counsel had failed to justify admitting that opinion under Federal Rule of Evidence 702.
The Union Pacific planned to have George Page, Manager of Ergonomics for the Union Pacific, state conclusions on whether Huffman’s osteoarthritis was likely the result of his work. As questions concerning his opinions began, though, the district court sustained an objection to his causation testimony on the grounds that he had not been designated as an expert.
On appeal, there is no challenge to these twin rulings.
We now examine the evidence that was admitted from which jurors drew their conclusions about causation and the railroad’s liability.
*423 1. Evidence of the kind of work Huffman performed
Dr. Robert Andres, a consultant on ergonomics employed by Huffman for the trial, testified regarding Huffman’s activities. Dr. Andres’s knowledge was based on a 30 to 45 minute telephone conversation with Huffman and consideration of Huffman’s deposition. Dr. Andres listed activities that Huffman completed as a trainman: he put on end-of-train devices, engaged in pin pulling, changed knuckles, and aligned drawbars. Dr. Andres also stated that these were the kinds of tasks he had observed completed by trainmen across the country. Huffman testified that he climbed on and off of moving equipment, set handbrakes, coupled air hoses, and walked the entire length of the train— 8,000 feet — on sloping ballast.
2. Evidence of the kind of injury Huffman suffered
It was uncontested that Huffman had osteoarthritis of the knees. Dr. Alan Smith, Huffman’s treating physician, testified regarding Huffman’s medical history. He had treated Huffman since 1992. In 1993, Huffman was diagnosed with patellar tendonitis, an inflammation of the tendon, and received two injections in his left knee. Dr. Smith testified that tendonitis is not the same as osteoarthritis. In September 1997, Huffman again saw Dr. Smith due to pain in his left knee. Dr. Smith believed the pain to be secondary to a “misuse injury” caused by Huffman’s walking differently after an ankle injury.
Dr. Smith stated that he first connected Huffman’s knee pain to osteoarthritis in December 2005, which was after Huffman’s retirement. In August 2007, Huffman received an injection in both knees as treatment for osteoarthritis. The injection was to reduce inflammation and to cushion the knee by increasing the amount and viscosity of the joint fluid. A year later, Huffman returned to Dr. Smith and complained of knee pain. Because the injections commonly last for about a year to a year-and-a-half, Huffman began a second round of injections in November 2008. Dr. Smith testified that if Huffman returns with knee pain within six months from the time of his last injection, he will recommend knee replacement surgery.
During oral argument, this court was informed that Huffinan will have a total knee replacement.
Based on the foregoing evidence, osteoarthritis is the only injury at issue in this case, because no evidence of any other injury was presented to the jury. Even though the jury instructions were broad and did not limit the injury to osteoarthritis, the jury could not consider an injury for which no evidence was produced. Further, the jury instructions with respect to the statute of limitations refer to the relevant injury as a knee-related injury.
3. Evidence of the kind of work trainmen perform
A document created by the Union Pacific identified the essential job functions of a trainman. The document listed coupling and uncoupling air hoses and electrical connections between cars, assisting with removal and replacement of broken knuckles, applying and releasing handbrakes, assisting in the alignment of drawbars, and getting on and off of stationary and moving equipment at less than three miles per hour. This is consistent with the testimony of Dr. Andres and Huffman on Huffman’s duties as a trainman.
A Evidence of the injuries that result from trainman work
Evidence was presented to demonstrate that the kind of work trainmen do can *424cause injury if not completed in the ergonomically optimal way.
Page, Manager of Ergonomics for the Union Pacific, discussed his research on how railroad employees operate equipment and the role of ergonomics. Page described ergonomics as having a safety component that asks, “how much a human can do in the workplace before they become at an elevated risk for ... lower extremity injuries of the leg.” He also discussed a video created to teach railroad employees the best method for lifting and how an inappropriate method of lifting can magnify stress on the lower back. Page stated that knee osteoarthritis and musculoskeletal disorders in trainmen were not on the Union Pacific’s priority list of ergonomic concerns.
Dr. Andres testified regarding ergonomic risk factors present in trainman activities. Dr. Andres read from a detailed 34-page report he had composed detailing Huffman’s exposure to ergonomic risk factors while working for the railroad and the effectiveness of the railroad’s efforts to combat musculoskeletal disorders. The force created by an activity and the worker’s posture at the time were said to be key to the possibility of injury. Dr. Andres identified squatting, kneeling, lifting, carrying, climbing or walking on uneven surfaces as ergonomic risk factors. He stated that these risk factors are present in trainman activities generally and specifically present in coupling air hoses and throwing switches.
Dr. Andres testified as to the desirability for a railroad to have a comprehensive ergonomics program to reduce injuries. His testimony tied the kind of work Huffman did to ergonomic risks to the lower limbs. He did not testify, though, that the risks included osteoarthritis in the knees.
There was evidence the Union Pacific gave some workers training about “ergonomics awareness.” This training recognized that “many musculoskeletal symptoms have multiple causes and may be associated with work and non-work factors.” The factors listed in the training include intense, long-duration exposures to combinations of force and repetition, doing tasks beyond physical capabilities, and exposure to vibration above maximum health guidelines. To reduce the factors associated with musculoskeletal symptoms, the training suggested using sound work practices like neutral body postures, keeping items within easy reach, getting help if a task requires overexertion, taking stretch breaks, and using safe lifting practices and good body mechanics. Huffman did not receive this training. None of this evidence established that failure to follow best ergonomic practices would lead to osteoarthritis in the knees.
Dr. Richard William Bunch, a licensed physical therapist, consulted with the Union Pacific on training programs for workers. Dr. Bunch defined musculoskeletal disorders as including “a whole array of problems that involve muscle, tendon, bone, ligaments, joints, even nerves and blood vessels.” Specifically, Dr. Bunch identified all sprains and strains, damages to the ligaments and tendons or the muscles, carpal tunnel syndrome, bone spurs, and herniated disks in the neck and back as examples of musculoskeletal disorders. Dr. Bunch, however, did not identify osteoarthritis of the knees or an inflammation of the articular cartilage in a joint as musculoskeletal disorders.
William Eugene Roe, Chief Engineer of one of the Union Pacific regions, was in charge of the Bodies-in-Motion program for the Engineering Department. That program was designed to make the job more comfortable to the worker and counter effects of muscle tiredness. This train*425ing was not provided to workers like Huffman in the Transportation Department.
Somewhat relevant here is the expert testimony Dr. Smith was allowed to give that explained arthritis as an “inflammation of the joint.” Osteoarthritis “is inflammation of the articular cartilage at a joint,” i.e., the cartilage that is “the contact surface between two bones at a joint.” As to cause, there was testimony that osteoarthritis is “wear and tear” and is “caused by life.”
With this as the evidence, we now turn to whether the evidence presented was enough for the jury to grant Huffman relief.
D. Analysis
The jury “may draw reasonable inferences from the evidence,” and we may not substitute other inferences for the “jury’s reasonable factual inferences.” Brown v. Parker Drilling Offshore Corp., 410 F.3d 166, 182-83 (5th Cir.2005) (quotation marks and citation omitted). In deciding whether Huffman offered enough evidence for the jury to grant him relief, we remind ourselves that “very little evidence is sufficient to make an issue one for the jury.” 9B Charles Alan Wright & Arthur R. Miller, Federal Practice and Procedure § 2526 (3d ed. 2008). We review this evidence in the light most favorable to the verdict. Rivera, 378 F.3d at 508-09. “Only when there is a complete absence of probative facts to support the conclusion reached does a reversible error appear.” Lavender, 327 U.S. at 653, 66 S.Ct. 740. As recently as in Rivera, we applied the “complete absence of probative facts” standard. Rivera, 378 F.3d at 505. We have also noted that “the Supreme Court used the term ‘slightest’ to describe the reduced standard of causation between the employer’s negligence and the employee’s injury in FELA § 51 cases.” Gautreaux, 107 F.3d at 335.
We emphasize, though, what this standard is not. Almost every jury will have been presented evidence. It is the complete absence of “probative” evidence on some element of the claim that will require us to deny recovery. Under this standard, there must be evidence to support that work the claimant performed led to the specific condition the claimant suffered — not a lot of evidence, not necessarily expert evidence, but something probative that supplies jurors with everything they need to which inferences can then be applied. It was necessary, then, for probative evidence to be introduced that work such as Huffman performed would play at least a small part in bringing about Huffman’s osteoarthritis.
There is no viable challenge to the evidence as to Huffman’s actual work and Huffman’s actual medical condition. Between his own testimony and that of various other witnesses, jurors would have known what he did on the railroad and that he had osteoarthritis in his knees.
On the other hand, no evidence was presented that the osteoarthritis Huffman had was a kind of musculoskeletal disorder that could occur if a railroad negligently failed to inform its trainmen how to perform their tasks. As previously discussed, we are only concerned with osteoarthritis because this is the only medical condition addressed by the evidence.
At best, there was evidence that the kind of work trainmen did, if not performed properly, could increase the chances of musculoskeletal disorders. The term musculoskeletal disorders, however, encompasses “a whole array of problems that involve muscle, tendon, bone, ligaments, joints, even nerves and blood vessels.” There was not any testimony listing *426osteoarthritis in the knees as a musculoskeletal disorder that could result from performing trainmen activities. There were some extremely limited references to stresses on the lower extremities resulting from such work but no link to osteoarthritis of the knees was made. Dr. Andres presented ergonomic risk factors that could lead to disorders of the lower extremities, but never tied those risk factors to osteoarthritis in the knees, or arthritis of any sort. Musculoskeletal disorder is too broad a category, and the evidence introduced too general, for jurors to have a basis on which to infer even the minimal degree of causation required.
We wish to be clear about what is missing. Jurors have wide latitude in FELA cases, and the quantum of causation that is required is low. Jurors still may not simply guess. Evidence that work performed by trainmen increased the risk of musculoskeletal disorders if not performed properly never identified osteoarthritis in the knees as one of those disorders that could result. The path from worker injury to employer liability was too broken in this record to allow juror common sense to travel it. See Fontenot, 714 F.2d at 20.
We hold that the evidence was insufficient on causation. What form the additional necessary evidence should have taken is not before us. Huffman had planned on including expert testimony on causation. We are not reviewing the sufficiency of what Huffman planned to introduce, but what he actually did. It was not enough.
We REVERSE and REMAND for entry of judgment for the Union Pacific.
. The term “cumulative trauma injuries” is used somewhat interchangeably in the record with the term "musculoskeletal disorders.”
. For example, one video discussed the hazards of walking on ballast, particularly sloped ballast, and gave advice as to the least physically stressful way to do so.
. Gautreaux involved the Jones Act, but we noted that "seamen are afforded rights parallei to those of railway employees under” FELA. 107 F.3d at 335.
. We were, post-argument, urged to consider Brooks v. Union Pac. R.R. Co., 620 F.3d 896 (8th Cir.2010). Brooks had been diagnosed with degenerative disc disease but claimed FELA liability due to an acute injury that occurred one day at work. Id. at 897-98. He had no evidence of the kind of work he usually performed and of the kinds of injuries such activities generally caused. Id. at 899. The court held, citing Claar, that because Brooks’ injury "had no obvious origin,” expert testimony was required. Id. We do not find Brooks and our case to be similar.