Opinion by
KAREN ANGELINI, Justice.This is a Federal Employers’ Liability Act (FELA) case in which Cenobio E. Navarro, representative of the Estate of Manuela Navarro, claims the exposure to diesel exhaust at the Union Pacific rail yard in Laredo, Texas, caused Manuela Navarro’s bone marrow cancer and resulting death. Following a jury trial, the trial court rendered judgment on the jury’s verdict in Navarro’s favor and awarded $2 million in damages. Missouri Pacific Railroad Company d/b/a Union Pacific Railroad Company (“Union Pacific”) appeals on two grounds: (1) that the trial court erred in admitting the testimony of Navarro’s expert witnesses because it was unreliable and (2) the evidence is legally insufficient to sustain the jury’s verdict in Navarro’s favor. We agree with Union Pacific and reverse the trial court’s judgment and render judgment in Union Pacific’s favor.
Factual and Procedural Background
Manuela Navarro (“Manuela”) worked as a part-time clerk at the Union Pacific Railroad Company rail yard in Laredo, Texas from 1974 until 1994. During this period of time, Manuela worked in various capacities and locations in the rail yard in which she alleged she was exposed to diesel exhaust. Manuela became ill in 1994 and was diagnosed with multiple myeloma. Manuela sued Union Pacific, alleging the diesel exhaust she was exposed to while working at Union Pacific caused her disease. She died in 1999, while the trial was ongoing.
Before the case was tried, Union Pacific filed Motions to Exclude Expert Testimony. After conducting hearings, the trial court denied the motions, and the case proceeded to trial. Union Pacific continued to object to Navarro’s expert witnesses’ testimony on the grounds that such testimony was scientifically unreliable. The trial court overruled the objections, however, and Navarro’s experts’ testimony was heard and considered by the jury. The jury found in Navarro’s favor, and the trial court entered judgment in the amount of $2 million against Union Pacific. On appeal, Union Pacific contends that Navarro’s expert witness testimony was unreliable and should have been excluded. Thus, according to Union Pacific, as a matter of law, there was no admissible evidence to prove Manuela’s bone marrow cancer was caused by exposure to diesel exhaust. Navarro argues, on the other hand, that the trial court acted within her discretion in admitting his expert witness’s testimony *750and that he presented sufficient evidence under the standard imposed by the FELA.
Standard op Review-Expert Testimony
We review evidentiary rulings, including rulings on expert testimony, for an abuse of discretion. See E.I. du Pont de Nemours & Co. v. Robinson, 923 S.W.2d 549, 558 (Tex.1995). Where expert testimony is not grounded in the methods and procedures of science, it amounts to nothing more than subjective belief or unsupported speculation. Id. at 557. In Robinson, the Texas Supreme Court adopted the test set forth by the United States Supreme Court in Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993), which held that in order to be admissible, expert testimony must be relevant and reliable. In Robinson, the supreme court set forth a non-exclusive list of factors for the courts, acting as gatekeeper, to consider in determining the reliability of expert testimony, as follows:
(1) the extent to which the theory has been or can be tested;
(2) the extent to which the technique relies upon the subjective interpretation of the expert;
(3) whether the theory has been subjected to peer review and/or publication;
(4) the technique’s potential rate of error;
(5) whether the underlying theory or technique has been generally accepted as valid by the relevant scientific community; and
(6) the non-judicial uses which have been made of the theory or technique.
Id. at 557.
The supreme court, recognizing that the Robinson-factors may not apply to non-scientific experts who base their opinions on individual experience, adopted the “analytical gap” analysis in Gammill v. Jack Williams Chevrolet, Inc., 972 S.W.2d 713, 727 (Tex.1998). Under this analysis, the expert may testify to his observations and experience, but still must show some basis for the opinion to demonstrate its reliability. In performing its gatekeeping role, the court “may conclude that there is simply too great an analytical gap between the data and the opinion proffered.” Id. at 726.
Where the trial court has admitted the expert testimony and the appellant challenges, on appeal, the expert testimony as constituting “no evidence,” we consider whether the expert testimony is reliable under a de novo standard of review. See Merrell Dow Pharms., Inc. v. Havner, 953 S.W.2d 706, 710-720 (Tex.1997); Austin v. Kerr-McGee Refining Corp., 25 S.W.3d 280, 285 (Tex.App.-Texarkana 2000, no pet.); Minnesota Mining & Mfg. Co. v. Atterbury, 978 S.W.2d 183, 192 (Tex.App.-Texarkana 1998, pet. denied); see generally, Ricky J. Poole and Kimberly S. Keller, Jury Erosion: The Effects of Robinson, Havner, & Gammill on the Role of Texas Juries, 32 St. Mary’s L.J. 383,418-420 (2001) (discussing de novo standard of review when trial court admits expert testimony). And, where the “no evidence” challenge is brought against causation experts, the proponent of the evidence must exclude other plausible causes of the injury with reasonable certainty. Havner, 953 S.W.2d at 720.
In cases brought under the FELA, the plaintiffs burden of proof has been referred to as a “featherweight” burden. See Rogers v. Missouri Pacific R.R. Co., 352 U.S. 500, 506-07, 77 S.Ct. 443, 1 L.Ed.2d 493 (1957); Maritime Overseas Corp. v. Ellis, 971 S.W.2d 402, 406 (Tex.1998). Under the FELA, the test of a jury case is simply “whether the proof justifies with reason the conclusion that *751employer negligence played any part, even the slightest, in producing the injury or death for which the claimant seeks damages.” Ellis, 971 S.W.2d at 406. This lower burden under the FELA, however, has not been generally applied to the admissibility of expert testimony. See Claar v. Burlington N. R.R. Co., 29 F.3d 499, 503 (9th Cir.1994); In re Paoli R.R. Yard PCB Litig., Nos. 86-2229, 87-1190, 87-1258, 87-3227, 2000 WL 274262 *2 (E.D.Pa. March 7, 2000); Savage v. Union Pacific R.R. Co., 67 F.Supp.2d 1021, 1029 (W.D.Ark.1999). In an FELA case, the Daubert standard of admissibility of expert evidence “extends to each step in an expert’s analysis all the way through the step that connects the work of the expert to the particular case.” In re Paoli R.R. Yard PCB Litig., 35 F.3d 717, 743 (3rd Cir.1994).
The Expert Testimony Evidence
Navarro presented the testimony of four expert witnesses to establish that exposure to diesel exhaust at the Laredo Union Pacific rail yard caused Manuela’s multiple myeloma and, ultimately, her death.
1. Frank M. Parker, III
Frank M. Parker, III, an industrial hygienist, testified that Manuela was exposed to significant levels of diesel exhaust at the Union Pacific rail yard. No air sampling was done at the rail yard during the time Manuela was working there. And, Parker did not take ah’ samples at any time. Parker estimated Manuela’s exposure based on a comparison of her description of her job duties to a single study— the Hammond study — conducted in the northeastern part of the United States. That study categorized rail yard workers by job classification and listed their respective levels of exposure. Using this study, Parker estimated Manuela’s exposure to diesel exhaust ranged between 128 and 233 micrograms per cubic meter. Although Manuela was classified as a clerk while working at the rail yard, Parker used the “freight braker” and “hostler” categories for comparison because he felt her duties were more consistent with those categories than with the clerk category. The level of exposure for the clerk category is 130 micrograms per cubic meter.
Parker testified the Environmental Protection Agency’s (“EPA”) odor threshold for diesel exhaust is 200 micrograms per cubic meter. However, the document containing this information was a draft document with a “do not cite or quote” notation. According to the EPA document, diesel exhaust is classified as a suspect human carcinogen and it is estimated that the average railroad worker is exposed to 125 micrograms per cubic meter of diesel exhaust.
Parker admitted his only professional experience relating to diesel exhaust exposure has been as an expert witness in this lawsuit. He also admitted his initial opinion was that Manuela had been exposed to only 7.2 micrograms per cubic meter of diesel exhaust. He reached this initial number by looking at the 130 level for clerks in the study and then adjusting that number by removing cigarette smoke, inorganic dust and other materials. That number is called the adjustable extractable mean as opposed to the respirable particulate mean, which is not adjusted for other materials. The adjustable extractable mean for brakers would be 7.9 rather than 128 and for hostlers it would be 33 rather than 233. Parker later changed his initial opinion on exposure because he ultimately considered Manuela’s job duties to be more like a hostler or braker than a clerk and he decided to use the respirable particulate mean measure rather than the ad*752justable extractable mean. By using a different job classification and by using the respirable particulate mean measure, his opinion then changed from an exposure of 7.2 micrograms per cubic meter to between 128 and 233 micrograms per cubic meter.
Parker did not allow for the variables the Hammond article warned against. According to the article, exposure amounts found in the study conducted in the northeastern part of the country did not necessarily reflect exposure amounts in other parts of the United States. The article pointed out that exposure rates are lower in warmer climates, that part-time employees would have different exposure amounts, and that diesel engines used after the 1960’s emit less diesel exhaust.
Parker admitted Manuela’s exposure may have been significantly higher or lower than his estimate. He did not try to actually quantify the degree of her exposure because the underlying assumptions he would have to make are shaky. Parker did not look at diesel exhaust components as part of evaluating levels of exposure in this case.
2. Dr. Hari Dayal
Dr. Hari Dayal, an expert in cancer epidemiology, also testified for Navarro. Dr. Dayal testified that railroad workers who come in contact with diesel exhaust have six times the risk of developing multiple myeloma. He also testified that exposure to diesel exhaust contributed to the development of Manuela’s multiple myelo-ma. In reaching his opinion, he looked at many studies. Ultimately, he based his opinion on only two studies. One study he relied on is the Boffetta study, conducted by the American Cancer Society, which, according to Dr. Dayal, showed that railroad workers have a six times greater risk of developing multiple myeloma. The Bof-fetta study, however, does not reach the same conclusion Dr. Dayal reaches. In the Boffetta study, there were only three cases of multiple myeloma in railroad workers reported. Thus, the authors of the Boffetta study concluded that the results regarding railroad workers (and several other occupations) and multiple mye-loma were inconclusive because they were either statistically insignificant or based on only a small number of exposed subjects. The authors concluded that people who reported they had been exposed to diesel exhaust had no statistically significant increased risk of getting multiple myeloma. The Boffetta study does show diabetes as a risk factor for multiple myeloma. It is undisputed that Manuela and others in her family suffered from diabetes.
Dr. Dayal also relied on the Hansen study. According to Dr. Dayal, the Hansen study shows a statistically significant relationship between being a truck driver extensively exposed to diesel exhaust and developing multiple myeloma. Dr. Dayal conceded that the author of the Hansen study actually concludes there is no relationship, but he believes that finding is a false negative. In other words, even though the study says there is no relationship, he believes there really is a relationship because the measurement of exposure was not perfect.
Dr. Hansen, the author of the Hansen study, admits there are problems with the data in her study. She concluded that, based on the observation of only five deaths from multiple myeloma, the finding was statistically significant but may have been due to chance. Dr. Dayal, however, does not agree with Dr. Hansen’s conclusion.
Dr. Dayal referred to some studies other than the Boffetta and Hansen studies to support his belief that there is a relationship between diesel exhaust exposure and *753multiple myeloma; however, these studies did not show a statistical significance. Dr. Dayal testified that if an epidemiological study finds a relationship between an exposure and a disease, you still must apply the Bradford Hill nine-step criteria. The Boffetta and Hansen studies meet the Bradford Hill criteria, but he did not review the other studies under this analysis.
Dr. Dayal testified he had never published any papers or articles relating to multiple myeloma, diesel exhaust exposure, or railroad workers. This is the first case he has worked on as an expert witness involving diesel exposure or railroad workers.
3. Dr. Frank Gardner
Dr. Frank Gardner also testified for Navarro. He is a medical doctor specializing in hematology and oncology. It is his opinion that exposure to diesel exhaust can cause multiple myeloma and that such exposure caused Manuela’s multiple myelo-ma. He is familiar with literature that recognizes that diesel exhaust contains carcinogens. Dr. Gardner was unable to testify that a specific diesel exhaust concentration or length of time of exposure causes multiple myeloma. He testified there is no quantification. Dr. Gardner has conducted no research regarding the carcinogenity of diesel exhaust — he has only done literature searches. Dr. Gardner advanced a theory that diesel exhaust could get into bone marrow by breathing, swallowing or absorbing it through the skin.
4. Dr. Marvin Legator
Dr. Marvin Legator, an expert in genetic toxicology and chemical carcinogens, testified there is no safe level of diesel exhaust. He testified he relied on the Hansen study. It is his opinion that if a person has been exposed to a carcinogenic substance and then develops a neoplasm, then in all probability that exposure caused or significantly contributed to the neoplasm. Manuela’s exposure to diesel exhaust, according to Dr. Legator, in all probability caused or significantly contributed to her multiple myeloma.
5.Dr. Otto Wong
Dr. Otto Wong, Union Pacific’s primary expert witness, is an occupational and environmental epidemiology researcher. In his opinion, there is no causal relationship between exposure to diesel exhaust and the risk of developing multiple myeloma. He performed his own study and he also looked at the Boffetta and Hansen studies, as well as other studies to support his opinion. The Boffetta study found only half of the railroad workers in the study were even exposed to diesel exhaust and concluded the risk factor for railroad workers was not statistically significant or it was based on too small a number of exposed subjects. Dr. Wong also reviewed the Hansen study and determined it showed a significant risk, but the numbers are too small to rely on. He reviewed other studies that show no increased risk. The majority of the studies show no statistically significant risk. Dr. Wong testified that in this case there has been no quantification of the exposure levels of a particular chemical agent. Dr. Wong did state that when there are no records of exposure levels, the best we can do is look at individual job titles and consider the exposure levels for those respective job titles.
Dr. Wong relied on 18 studies for his opinion. He believes that one should consider a number of studies, including some that are large and that it is unfair to rely on just one or two studies. He agreed that some publications show diesel exhaust as a possible carcinogen, but they only refer to lung and bladder cancer — not multiple myeloma.
*754Discussion
This is a toxic tort case dependent upon epidemiological expert testimony. In deciding this case, we are guided by the Texas Supreme Court cases of Robinson, Havner and Gammill. In Robinson, the issue was admissibility of the evidence and, in making its determination, the court set forth the six non-exclusive factors previously listed. See Robinson, 923 S.W.2d at 557. Havner involved a no evidence review of scientific factors, but the supreme court applied the same factors. See Havner, 953 S.W.2d at 714.
The supreme court in Havner concluded:
If the foundational data underlying opinion testimony are unreliable, an expert will not be permitted to base an opinion on that data because any opinion drawn from that data is likewise unreliable. Further, an expert’s testimony is unreliable even when the underlying data are sound if the expert draws conclusions from that data based on flawed methodology. A flaw in the expert’s reasoning from the data may render rebanee on a study unreasonable and render the inferences drawn therefrom dubious. Under that circumstance, the expert’s scientific testimony is unrehable and, legally, no evidence.
Havner, 953 S.W.2d at 714.
In Havner, the supreme court also discussed the plaintiffs burden to prove both general causation and specific causation in a toxic tort case. See id. “General causation is whether a substance is capable of causing a particular injury or condition in the general population, while specific causation is whether a substance caused a particular individual’s injury.” Id. Recognizing that direct experimentation may not be possible to prove causation, the supreme court noted that plaintiffs may try to demonstrate that exposure to the substance at issue increases the risk of the particular injury or disease at issue. ’“The finder of fact is asked to infer that because the risk is demonstrably greater in the general population due to exposure to the substance, the claimant’s injury was more bkely than not caused by that substance.” Id. at 715.
“Epidemiological studies examine existing populations to attempt to determine if there is an association between a disease or condition and a factor suspected of causing that disease or condition.” Id. After discussing various courts’ approaches to using epidemiological studies in determining whether a fact issue is raised as to causation, the supreme court concluded that “properly designed and executed epidemiological studies may be part of the evidence supporting causation in a toxic tort case and that there is a rational basis for relating the requirement that there be more than a ‘doubling of the risk’ to our no evidence standard of review and to the more bkely than not burden of proof.” Id. at 717.
The supreme court further stated that other factors must be considered because epidemiological studies show only an association. Id. at 718. Many experts use the Bradford Hill criteria. Id. n. 2. Further, the design and execution of epidemiological studies must be examined. Id. at 719. One reason for such scrutiny is that bias can “dramatically affect the scientific reliability of an epidemiological study.” Id. And, the supreme court noted, “an expert cannot dissect a study, picking and choosing data, or ‘reanalyze’ the data to derive a higher relative risk if this process does not comport with sound scientific methodology.” Id. at 720.
An injured person, when using epidemiological studies, must also prove that he or she was exposed to the same sub*755stance analyzed in the studies, that the exposure was comparable to or greater than those in the studies, that the exposure occurred before the onset of injury, and the timing of the onset of injury was consistent with that experienced by those in the studies. Id. If there are other plausible causes of the injury that could be negated, the plaintiff must also offer evidence excluding those causes with reasonable certainty. Id. Concluding, the supreme court stated:
[W]e emphasize that courts must make a determination of reliability from all the evidence. Courts should allow a party, plaintiff or defendant, to present the best available evidence, assuming it passes muster under Robinson, and only then should a court determine from a totality of the evidence, considering all factors affecting the reliability of particular studies, whether there is legally sufficient evidence to support a judgment.
Id. at 720.
The supreme court in Havner reiterated that even though there appears to be an increased risk associated with an activity, this does not prove a causal relationship. Id. at 724. Other factors to consider in evaluating the reliability of a scientific study include: “the sample size of the study, the power of the study, confounding variables, and whether there was selection bias.” Id.
In Havner, the supreme court stressed the importance of publication and peer review. In Havner, none of the plaintiffs’ experts’ findings had been published, studied or replicated by the scientific community. In fact, the only review of their work was by judges and juries and the only publication was in the pages of federal and state reporters. Further, the studies were prepared solely for litigation. Id. at 726. Although publication is not a prerequisite for scientific reliability in every case, the supreme court stated courts must be particularly skeptical of scientific evidence that has not been published or subjected to peer review. Id. at 727. The supreme court went on to say that an isolated study finding a statistically significant association between the substance at issue and the condition at issue would not be legally sufficient evidence of causation. Id. “[W]hen a number of studies have been done, it would not be good practice to pick out one to support a conclusion.” Id.
Following the dictates of the supreme court to determine whether Navarro’s expert witnesses’ testimony is scientifically reliable, we consider whether there is evidence that exposure to diesel exhaust, in general, causes multiple myeloma. Havner, 953 S.W.2d at 714-18; Austin, 25 S.W.3d at 292. If so, then we ask whether there is evidence that exposure to diesel exhaust in this case caused Manuela’s multiple myeloma. In making this determination, we must first consider whether there is evidence of the amount of diesel exhaust Manuela was exposed to so that her exposure can be compared to those in the studies her experts rely upon. Havner, 953 S.W.2d at 720; Austin, 25 S.W.3d at 292.
We turn first to Frank Parker’s testimony which Navarro offered for the purpose of establishing the level of exposure to Manuela at the Laredo rail yard.1 It is undisputed that Navarro, as plaintiff in a toxic tort case, must show that Manuela was exposed to diesel exhaust and that her exposure was equal to or greater than *756the exposure in the studies upon which he relies. See Austin, 25 S.W.3d at 292.
A review of Frank Parker’s testimony reveals serious flaws iñ his methodology. First, Parker relied, principally, on only one study — the 1988 Hammond study of railroad workers exposed to diesel exhaust in the northeastern United States — to estimate Manuela’s level of exposure. Navarro points out that the Union Pacific expert witness, Dr. Wong, testified approvingly of the use of estimates such as Parker used here. While it is true that Parker was unable to calculate Manuela’s actual exposure because of the lack of tésting at the Laredo rail yard during the time in question, there was other data available from which estimates could be made. There was testing done in the Laredo rail yard beginning in 1999 and there was data from other train yards in the United States. Parker, however, chose to rely on the Hammond study only.
However, even if it were acceptable for Parker to rely only on the Hammond study to the exclusion of other available data, his analysis is further flawed. Manuela held a clerk position while working for Union Pacific. Parker’s original estimate of Manuela’s exposure, based on the Hammond study’s clerk classification, was 7.2 micrograms per cubic meter. He later changed this estimate significantly to between 128 and 233 micrograms per cubic meter. He obtained these higher numbers by comparing Manuela’s duties to a “hostler” or “braker” because he felt they more accurately fit her job duties. Further, he changed his analysis to use the respirable particulate mass measure rather than the adjustable extractable mass measure. In other words, he used a measurement that included tobacco smoke, dust, sand and other particles rather than using the more accurate measure of diesel exhaust particles in the air which he had initially used. Navarro refers to this shift in Parker’s opinion as an expert “refining calculations and perfecting reports through the time of trial.” See Foust v. Estate of Walters, 21 S.W.3d 495, 504 (Tex.App.-San Antonio 2000, pet. denied) (citing Exxon Corp. v. West Texas Gathering Co., 868 S.W.2d 299, 304 (Tex.1993)). Such a radical upward adjustment (from 7.2 to between 128 and 233), however, is more than a refinement; instead, it appears to be “devised to support a causation opinion without rehable bases to do so.” See Castellow v. Chevron U.S.A, 97 F.Supp.2d 780, 797 (S.D.Texas 2000).
Furthermore, although relying on the Hammond study, Parker faded to address the caution expressed by Hammond to others using her data. Specifically, Hammond cautioned that exposure rates in the Northeast railroad study do not reflect exposure rates in other parts of the United States.
As further basis for his opinion, Parker relied on an EPA proposed draft bulletin which indicated that the level at which most people can smell diesel exhaust is 200 micrograms per cubic meter. Because Manuela reported she smelled diesel exhaust at work, Parker opined she was exposed to at least 200 micrograms per cubic meter. The EPA draft bulletin, however, was marked “DRAFT — DO NOT CITE OR QUOTE,” which seems to belie the accuracy and reliability of the statement made in the document.
Navarro points out Parker’s undisputed extensive qualifications and credentials as an industrial hygienist. Parker testified, however, that his only professional experience with diesel exhaust exposure has been as an expert witness in this case.
For all the above-stated reasons, we find Parker’s testimony to be unreliable and, therefore, it should not have been *757admitted. Absent Parker’s estimates of Manuela’s diesel exhaust exposure, the other three expert witnesses’s testimony cannot stand because they each relied on Parker’s exposure estimate. However, even if we were to assume Parker’s estimates were based on sound scientific methodology, we find that the opinions of the three causation expert’s are likewise based on flawed methodology.
Dr. Dayal’s opinion that Manuela’s multiple myeloma was caused by exposure to diesel exhaust was based on two studies— the Boffetta study and the Hansen study. The author of the Boffetta study noted that because there were only three railroad worker deaths from multiple myeloma in the study, there was no statistical significance between railroad workers and multiple myeloma. Further, the Boffetta study recognized that less than 50% of railroad workers even reported exposure to diesel exhaust. Thus, it is impossible to tell whether any or all of the three railroad workers who developed multiple myeloma were even exposed to diesel exhaust. Dr. Dayal disagreed with Boffetta’s refusal to find a statistically significant relationship. In other words, Dr. Dayal took the data collected by Boffetta and reached a conclusion regarding causation that Boffetta himself was unwilling to make.
The Hansen study is similar to the Boffetta study in that the author likewise refused to find that exposure to diesel exhaust causes multiple myeloma. Dr. Hansen concluded that, although the finding was statistically significant, it may have been due to chance. Dr. Dayal disagreed and chose to rely on the Hansen report for his opinion that exposure to diesel exhaust causes multiple myeloma. Dr. Dayal’s methodology for determining causation has been rejected as unreliable. See General Electric Co. v. Joiner, 522 U.S. 136, 145, 118 S.Ct. 512, 139 L.Ed.2d 508 (1997) (holding where authors of study on which plaintiffs experts relied were unwilling to find causal relationship between exposure to toxic substance and plaintiffs disease, such study does not support plaintiffs expert’s causation conclusion).
Dr. Dayal also looked at several other studies, some of which he stated were almost statistically significant. And, he admitted there were other studies which showed no association between diesel exhaust and multiple myeloma. He chose to ignore those studies, however, and only relied on his interpretation of the data collected by Boffetta and Hansen. Dr. Dayal’s testimony is further flawed in that he was unable to testify as to what level of diesel exhaust would create a risk of multiple myeloma. See National Bank of Commerce v. Associated Milk Producers, 22 F.Supp.2d 942, 983 (N.D.Ark.1998); Sutera v. The Perrier Group of Amenca, 986 F.Supp. 655, 664 (D.Mass.1997).
Dr. Dayal’s methodology is rendered further suspect because of his total lack of experience in dealing with diesel exhaust and multiple myeloma. He has not conducted research or published papers on the subject.
Dr. Gardner testified that diesel exhaust can cause multiple myeloma and that it did cause Manuela’s multiple myeloma. As with Dr. Dayal, Dr. Gardner’s testimony is flawed in that he could not state what level of diesel exhaust is necessary for multiple myeloma to develop. Robinson, 923 S.W.2d at 559. As stated by Dr. Gardner, there is no quantification. When referring to diesel exhaust as a carcinogen, Dr. Gardner only testified that exposure causes cancer — he did not refer to a particular type of cancer other than lung cancer.
Dr. Gardner did suggest that diesel exhaust could be breathed, swallowed, or *758absorbed through the skin and then get into the bone marrow. He offered no support for this theory, however. -Dr. Gardner’s testimony was scientifically unreliable and should have been excluded.
Dr. Legator testified that there is no safe level of exposure to certain components of diesel exhaust. This type of reasoning has been rejected. Sutera, 986 F.Supp. at 665-67; National Bank of Commerce v. Associated, Milk Producers, 22 F.Supp.2d at 946.
Navarro was required to show not only general causation, but also specific causation. Specific causation is generally shown by a doubling of the risk factor. Havner, 953 S.W.2d at 715. Navarro has not shown any studies where exposure to diesel exhaust caused the risk of developing multiple myeloma to double.
Additionally, under Hamer, if there are other plausible causes of the condition, they should be negated. Navarro was required to exclude those causes with reasonable certainty. Havner, 953 S.W.2d at 720. It is undisputed that Manuela suffered from diabetes, had a family history of diabetes, and that the studies Navarro relied on found diabetes to be a risk factor for multiple myeloma. Yet, none of Navarro’s expert witnesses ruled out diabetes as a causation factor.
Because this is a toxic tort case, it is appropriate to consider the Robinson factors, in addition to the general rules established by Havner, in determining the reliability of the expert testimony. See Robinson, 923 S.W.2d at 557. Although the Robinson factors are non-exclusive, their application to the expert testimony in this case reveals Navarro’s experts’ flawed methodology.
The theory that diesel exhaust causes multiple myeloma has been tested in the many epidemiological studies that were reviewed in this case by the experts. None of the studies found a causal relationship— Navarro’s experts are the only ones who have been willing to make the connection, based on only two studies, whose authors expressly declined to find causation. The theories Navarro’s experts relied on are quite subjective, given the fact that no scientific studies support their opinions. The theory that exposure to diesel exhaust causes multiple myeloma has been subjected to peer review and the majority found no heightened risk. The technique’s potential rate of error is incapable of being analyzed. The evidence clearly shows Navarro’s witnesses’ theories have not been generally accepted as valid by the scientific community and the theories have been put to no non-judicial uses. The record shows that Navarro’s expert witnesses are alone in the scientific community in their opinions that exposure to diesel exhaust causes multiple myeloma. Their opinions were developed solely for the purpose of using them to prove causation in this lawsuit. Opinions that have been formed only for the purpose of testifying are more likely to be biased toward a particular result. Robinson, 923 S.W.2d at 559.
Even under the reduced burden of the “analytical gap” analysis, Navarro’s expert witness testimony is flawed. See Gammill, 972 S.W.2d at 727 (applying “analytical gap” theory to expert witness testimony). For each of Navarro’s four expert witnesses, one must make a huge leap from the data in the studies relied upon to arrive at the conclusion that exposure to diesel exhaust causes multiple myeloma and further, that such exposure caused Manuela’s multiple myeloma. We conclude “there is simply too great an analytical gap between the data and opinions proffered.” Id. at 726.
Conclusion
Upon a review of the evidence, we find that the expert testimony offered by Na*759varro to prove exposure to diesel exhaust causes multiple myeloma to be unreliable. Thus, there is no evidence to support the jury verdict. We sustain Union Pacific’s two issues on appeal and reverse the trial court’s judgment and render judgment in favor of Union Pacific.
. Navarro’s other three experts rely on Parker's exposure estimates to support their opinions regarding causation.