dissenting.
Reynolds claims that job-related mental stress and worry caused him to suffer a stroke. Reynolds was an officer of Reynolds Construction Company, a family-operated business. He testified before the ALJ that, as he left a construction bid meeting at work, he became sick with a severe headache, sweating, and cold chills in part of his head. He went home early but became even sicker and eventually went to the hospital that night. Medical evidence presented by the employer showed that he was finally diagnosed as having suffered a cerebrovascular accident involving an occlusion of the right vertebral artery, commonly called a stroke.
Reynolds, who proceeded pro se at the hearing before the ALJ, testified that, because of the company’s continuing financial problems, he had been working long hours and had been experiencing continuous job-related stress for about three years prior to the stroke. Reynolds did not claim that physical fatigue or exertion from working long hours caused or contributed to the stroke. Rather, he claimed that long hours of continuous mental stress and worry on the job caused or contributed to the stroke. Reynolds’ brother, also an officer of the company, also testified that Reynolds had been experiencing mental stress and pressure because of the company’s financial problems. He stated: “[W]e were all under duress, pressure, and strain and stress . . . [The claimant] had a lot of pressure on him. I can go get a fishing pole and leave mine alone, but he couldn’t do it.”
The only other evidence presented was the medical records of Reynolds’ physician. It was the physician’s opinion that Reynolds’ stroke was not job-related but was the result of his arteriosclerosis. Other medical evidence showed that the stroke was secondary to Reynolds’ existing diabetes, hypertension and hypertriglyceridemia.
In concluding that the stroke arose out of and in the course of Reynolds’ employment, the ALJ found that, “[f]rom the lay observations and opinions given by employer [sic] Reynolds and his brother and from the natural inferences and through human experience of this finder of fact, it is found that the employee’s stroke and the occlusion of the posterior inferior cerebellar artery was triggered by the extremely stressful conditions and long hours the employee was ex*28periencing during the three year period prior to the stroke.”
The employer contends that the ALJ, as affirmed by the appellate division and the superior court, erred in using the “natural inference through human experience” as the basis for determining that the stroke was a compensable injury, where there was uncontradicted medical evidence showing that the stroke was not work-related. Although I agree that the ALJ and the appellate division erred by using the “natural inference through human experience” as the basis for finding that the stroke was a compensable injury, I do so for different reasons than those urged by the employer.
The employer’s contentions with respect to the application of the “natural inference” rule and OCGA § 34-9-1 (4) are misplaced.
First, the Supreme Court has determined that, where the “natural inference” utilized by the ALJ in this case is available, it is competent and credible evidence of causation in cases where there is no medical evidence of causation, and that, even where medical evidence as to causation is presented, the “natural inference” does not disappear. Guye v. Home Indem. Co., 241 Ga. 213, 215-217 (244 SE2d 864) (1978); Southwire Co. v. Cato, 250 Ga. 895, 897-898 (302 SE2d 91) (1983); LaVista Equip. Supply v. Elliott, 186 Ga. App. 585, 587 (367 SE2d 811) (1988).
Secondly, as to the employer’s argument that the evidence failed to satisfy the preponderance of the evidence requirement of OCGA § 34-9-1 (4), this argument is addressed to the trier of fact, not this Court, which is bound by the “any evidence” rule applicable to appellate review. Guye, supra at 215. Moreover, the preponderance of the evidence requirement of OCGA § 34-9-1 (4) applies to cases involving “heart disease, heart attack, the failure or occlusion of any of the coronary blood vessels, or thrombosis.” It does not apply when the injury at issue is a stroke involving a cerebrovascular occlusion. It may be argued that because the statutory list of medical conditions concludes with the general term “thrombosis,” the statute embraces conditions unrelated to the prior enumerated coronary problems and could be construed to include a thrombosis or occlusion of a cerebral vessel. However, I find no evidence that the Legislature intended this broad construction. “It is a well-recognized rule of construction that when a statute or document enumerates by name several particular things, and concludes with a general term of enlargement, this latter term is to be construed as being ejusdem generis with the things specifically named, unless, of course, there is something to show that a wider sense was intended.” (Citations and punctuation omitted.) Wester v. State, 205 Ga. App. 336 (422 SE2d 433) (1992). Because all of the preceding terms relate to coronary problems, I conclude that the term “thrombosis” in the statute refers to a coronary thrombosis.
Nevertheless, I conclude the ALJ and the appellate division erro*29neously based the award on the “natural inference through human experience” because the “natural inference” was not available in this case to supply competent evidence that Reynolds’ job-related mental stress and worry caused or contributed to his stroke.
I agree with the general statements in the majority opinion that the factfinder is entitled to rely on the “natural inference through human experience” to establish causation and that the factfinder may reject the opinions of medical experts. The Supreme Court has clearly held that, where the “natural inference” is available, it is competent evidence of causation sufficient to satisfy the “any evidence” rule and that the “natural inference” does not disappear when expert medical evidence as to causation is presented. Gaye, supra at 215.-217; Southwire Co., supra at 897-898. Nevertheless, the issue which the majority does not address is whether there are necessarily certain types of cases where the “natural inference” should not be considered competent evidence of causation and whether this is such a case. This Court and the Supreme Court have found the “natural inference” available in numerous cases involving simple medical causation issues where a cause and effect relationship between the work and the injury may be inferred based on common sense and human experience. However, it follows that the “natural inference” should not be available to the factfinder in cases where the issue of a causal relationship between the work and the injury involves more complicated medical questions beyond the ken of laypersons and which require expert medical testimony to establish causation. Placing limits on the availability of the “natural inference” in cases involving complex medical causation issues is the issue of first impression on which this case turns.
Reynolds had the burden of showing by competent evidence that his stroke arose out of and in the course of his employment. The disputed issue in this case is not whether the stroke occurred in the course of employment, but whether it arose out of employment. “The words ‘arising out of’ mean that there must be some causal connection between the conditions under which the employee worked and the injury which he received. The causative danger must be incidental to the character of the employment, and not independent of the relation of master and servant.” Thornton v. Hartford Accident &c. Co., 198 Ga. 786, 792 (32 SE2d 816) (1945). In other words, “it excludes an injury which can not fairly be traced to the employment as a contributing proximate cause, and which comes from a hazard to which the [employee] would have been equally exposed apart from the employment.” Fried v. United States Fidelity &c. Co., 192 Ga. 492, 495 (15 SE2d 704) (1941). The rule is also that an employer takes the employee as it finds him and assumes the risk that a requirement of the job may aggravate an existing disease or condition. Griggs v. Lumber*30men’s Mut. Cas. Co., 61 Ga. App. 448 (6 SE2d 180) (1939), aff’d, 190 Ga. 277 (9 SE2d 84) (1940). Here, the medical evidence showed that Reynolds suffered from existing conditions or diseases which increased the risk of a stroke. Accordingly, if a requirement of Reynolds’ job either directly caused or contributed to the stroke, or aggravated an existing condition or disease in such a manner as to cause or contribute to the stroke, then the stroke arose from the employment and would be compensable. On the other hand, if the stroke was solely the result of the natural progression of an existing disease or condition or some other factor unrelated to a condition of employment, then the stroke did not arise out of the employment and would not be compensable. Sutton v. B & L Express, 215 Ga. App. 394 (450 SE2d 859) (1994).
It has been recognized that the burden of proving causation under these standards may be difficult to carry in cases of this nature. For example, in heart attack cases “it is often difficult for the trier of fact to find the line between a noncompensable heart injury that is a symptom of an existing disease merely manifested during job exertion, and a compensable heart injury to which the job exertion was a contributing precipitating factor. [Cits.] The generalized and complex nature of a heart injury may prevent causation from being conclusively attributed to the work performed. In determining whether claimant’s job activities caused or contributed to his heart injury, it has been said that the factfinder may rely on different forms of evidence, including medical opinion or ‘the natural inference through human experience.’ Carter v. Kansas City Fire &c. Ins. Co., [138 Ga. App. 601, 604 (226 SE2d 755) (1976)].” Guye, supra at 215-216. The “natural inference” rule has been stated as follows: “ ‘[T]he evidence must show the work engaged in by the employer to have been sufficiently strenuous or of such a nature that, combined with the other facts of the case, it raises a natural inference through human experience that the exertion contributed toward the precipitation of the attack. . . .’ Hoffman v. Nat. Surety Corp., 91 Ga. App. 414 (85 SE2d 784) (1955).” Southwire Co., supra at 897. The same “natural inference” rule has been applied in cases involving cerebral vascular accident or stroke. Employers Mut. Liability Ins. Co. v. Videtto, 124 Ga. App,. 458, 461 (184 SE2d 210) (1971).3
In concluding that Reynolds’ job-related mental stress caused or contributed to his stroke, the ALJ relied on two types of evidence: (1) the lay observations and opinions of Reynolds and his brother, and (2) the natural inference through human experience. The award must *31be affirmed if there was any competent evidence of causation to support it. Because I conclude that neither lay observations and opinions nor the “natural inference” was competent evidence of causation in this case, there was a lack of any evidence of causation to support the award.
First, I address the ALJ’s use of the “natural inference” in this case to establish a causal connection between continuous job-related mental stress and worry and the stroke. Although the majority characterizes this as a case involving employment-related “stress and fatigue,” the record shows that Reynolds did not claim the stroke was caused by any type of physical exertion on the job. Rather, he claimed that the stroke was caused by three years of job-related mental stress and worry.
The primary use of the “natural inference” has been in cases where the claimant alleged that a heart attack was caused by job-related physical exertion. In Guye, supra, the Supreme Court determined that the “natural inference” was available as competent evidence that a heart attack was caused by strenuous manual labor at work. See also Carter, supra (natural inference available when heart attack followed physical exertion at work). In other heart attack cases, it has been determined that the “natural inference” was not available, either because the physical exertion was not sufficient to support the inference (see Hoffman, supra), or because facts otherwise failed to show a sufficient connection between the physical exertion at work and the onset of the attack. See Southwire Co., supra at 898 (natural inference not available where the symptoms of the heart attack did not occur until the claimant had been home for several hours); see also Travelers Ins. Co. v. Maddox, 118 Ga. App. 596, 598-599 (164 SE2d 850) (1968) (physical precedent) (Quillian, J., concurring specially on basis that the job-related physical exertion was not sufficient to raise a natural inference that it precipitated the heart attack); compare Gallman v. Coronet Indus., 182 Ga. App. 649, 650 (356 SE2d 654) (1987) (sufficient physical exertion demonstrated to authorize use of the natural inference). The “natural inference” has also been found available in cases where the claimant asserted that job-related physical exertion caused or contributed to a stroke. Videtto, supra; Hartford Accident & Indem. Co. v. Waters, 87 Ga. App. 117 (73 SE2d 70) (1952).
Although the “natural inference” has been applied in cases where mental stress, worry, or excitement has been claimed as contributing factors in a heart attack, in these cases there has also been evidence that physical exertion caused or contributed to the heart attack. See Southwire Co. v. Eason, 181 Ga. App. 708 (353 SE2d 567) (1987) (natural inference found available in case where evidence showed that the heart attack symptoms started during claimant’s physically stren*32uous workshift and claimant’s doctor testified that the work was also emotionally stressful); Sewell v. Bill Johnson Motors, 213 Ga. App. 853 (446 SE2d 239) (1994) (natural inference available where evidence showed claimant’s job was physically and emotionally demanding, that the heart attack symptoms started on the job following work-related exertion, and claimant was under unusually high work-related stress on the afternoon prior to the heart attack).
In cases where the claim was that mental or emotional stress alone caused a heart attack or heart disease, awards for the claimant have been affirmed based on expert medical testimony in support of causation without reliance on the “natural inference.” City Council of Augusta v. Williams, 137 Ga. App. 177 (223 SE2d 227) (1976); Zippy Mart v. Fender, 170 Ga. App. 617 (317 SE2d 575) (1984). The same rule has been applied in cases where the claimant contended that mental stress alone caused a stroke. In Ga. Bureau of Investigation v. Worthington, 149 Ga. App. 628 (255 SE2d 99) (1979), the claimant contended that job-related stress and tension caused his injury. See Williams v. ARA Environmental Svcs., 175 Ga. App. 661, 662 (334 SE2d 192) (1985) (noting that the record in Worthington showed a diagnosis of a stroke). An award for the claimant in Worthington was affirmed on the basis of expert medical testimony that the stress precipitated the stroke without reference to the “natural inference.” In H. M. Patterson & Sons v. Glover, 128 Ga. App. 736 (197 SE2d 851) (1973), the claimant contended that his light job duties as a hearse driver, a “traumatic experience” at work, and other job-related “consternation,” caused or contributed to his stroke. In reversing an award for the claimant which was unsupported by any medical evidence of causation, the Court held that the “natural inference” was not available to establish that the job conditions caused the stroke. Id. at 739.
As the above analysis demonstrates, the cases in which the “natural inference” has been relied upon to establish a causal connection between a condition of employment and an injury, including those where mental or emotional stress was a factor, have all involved evidence that some type of physical exertion caused or contributed to the injury. These are the type of cases where a cause and effect relationship between the strenuous work conditions and the injury can be inferred based on common sense and human experience. See OCGA §§ 24-1-1 (6); 24-4-9. Awards based solely on claims that job-related mental or emotional stress caused an injury have relied on expert medical evidence rather than the “natural inference” to establish causation.4 The “natural inference” has not been relied upon as compe*33tent evidence of causation in a case of the present type, where Reynolds claims that the cumulative effects of three years of job-related mental stress and worry caused him to suffer a stroke.
Two underlying reasons have been associated with application of the “natural inference through human experience” as competent evidence of causation in some cases, while in other cases expert medical evidence should be required to supply competent evidence of causation. Larson, 2B, The Law of Workmen’s Compensation, § 79.53. First, in. some cases lay testimony, including testimony from the claimant, has probative value in establishing simple matters such as the fact and location of pain, the nature of the events surrounding the injury, and the fact and extent of the claimant’s ability or inability to work. See Thomas v. United States Cas. Co., 218 Ga. 493 (128 SE2d 749) (1962); Fulton Bag & Cotton Mills v. Speaks, 90 Ga. App. 685, 692-696 (83 SE2d 872) (1954); Truelove v. Hulette, 103 Ga. App. 641 (120 SE2d 342) (1961); Kelley v. West Point Pepperell, 164 Ga. App. 187 (296 SE2d 191) (1982). Second, in some types of cases it may be possible to rely on a demonstrated expertise developed by the factfinder in analyzing uncomplicated kinds of medical facts, particularly in such matters as disability, malingering, and other issues not involving purely medical questions. Larson, supra at § 79.53; see Autry v. General Motors &c. Plant, 85 Ga. App. 500 (69 SE2d 697) (1952). On the other hand, “reliance on lay testimony and administrative expertise [as competent evidence of causation] is not justified when the medical question is no longer an uncomplicated one and carries the factfinders into realms that are properly within the province of medical experts.” (Emphasis omitted.) Larson, supra at § 79.54 (a).
Whether a causal connection exists between the continuous job-related mental stress experienced by Reynolds over a period of three years and his stroke involving a cerebrovascular occlusion is not a simple medical question about which a cause and effect relationship may be inferred by laymen based on common sense or human experience. It is a medical question of the sort that requires expert medical testimony. See Nat. Dairy Products Corp. v. Durham, 115 Ga. App. 420 (154 SE2d 752) (1967); Jordan v. United Ins. Co. &c., 158 Ga. App. 520, 521 (281 SE2d 286) (1981); see generally Cherokee County Hosp. Auth. v. Beaver, 179 Ga. App. 200, 204 (345 SE2d 904) (1986). As this Court recognized in setting forth the “natural inference” rule in Hoffman, supra, some cases are appropriate for use of the “natural inference” to establish causation, but “[i]n other cases, the opinions *34of experts that [a causal connection exists] is also sufficient. . . . [I]n one way or another, the fact must appear.” Id. at 417. And as stated in McDaniel v. Employers Mut. Liab. Ins. Co., 104 Ga. App. 340, 343 (121 SE2d 801) (1961), where the natural inference is not available in a case of this nature to establish causation, the issue of causation “becomes solely a medical question. . . .” See also Callaway Mills Co. v. Hurley, 100 Ga. App. 781 (112 SE2d 320) (1959). Accordingly, I conclude that the ALJ erred by using the “natural inference” in this case to establish the necessary causal connection between the stroke and the continuous job-related mental stress and worry experienced by Reynolds over a period of three years.
It follows from these conclusions that the second form of evidence relied upon by the ALJ, lay observations and opinions given by Reynolds and his brother, was not sufficient to establish causation in this case. The three forms of evidence which may be used by the factfinder to establish a causal connection between the job conditions and a resulting heart attack or similar type injury were identified in Carter, supra at 604, as follows: “medical opinion, lay observations and opinion, and ‘the natural inference through human experience.’ See McDaniel v. Employers Mut. Liab. Ins. Co., 104 Ga. App. 340 (2), supra. See also dissent in Brown Transport v. Jenkins, 129 Ga. App. 457, 463 (199 SE2d 910) [(1973)].” The supporting opinions in McDaniel, supra, and Brown Transport, supra, make clear that “lay observations and opinion” refer not to lay opinion as to causation, but to lay opinion and observations about the strenuous nature of the work, the symptoms of the injury, the disability experienced by the claimant, and other factual matters surrounding the work and the injury. Obviously, to accept lay opinion testimony as competent evidence of causation in a case of this nature would be to substitute another “natural inference” as to causation for the same “natural inference” not available to the factfinder.
Since neither the “natural inference through human experience” nor the lay opinions and observations presented by Reynolds were competent evidence of causation, and Reynolds produced no medical evidence in support of causation, there was no competent evidence of causation necessary to support the award, which must be reversed. This is not to say that the factfinder was required to accept the expert medical testimony presented by the employer. It is to say that Reynolds had the burden tb produce some competent evidence in support of causation and he produced none. Nevertheless, I would hold that, since the unavailability of the “natural inference” in this case was “an issue of first impression whose resolution was not clearly foreshadowed,” Reynolds should be given an opportunity to furnish additional competent evidence of causation to avoid imposing any “injustice or hardship.” Flewellen v. Atlanta Cas. Co., 250 Ga. 709, *35712 (300 SE2d 673) (1983). Accordingly, as in Hoffman, supra at 417, and Southwire Co. v. Cato, supra at 898, this case should be remanded for further proceedings before the ALJ for this purpose.
Decided July 14, 1995 Whelchel, Brown, Readdick & Bumgartner, Richard A. Brown, Jr., Gregory T. Carter,. Swift, Currie, McGhee & Hiers, James T. McDonald, Jr., for appellants. Edward E. Boshears, for appellee.This “natural inference” rule is not the same as the natural presumption applicable in unexplained death cases. See Southwire Co., supra at 896-897; LaVista Equip., supra at 587; Odom v. Transamerica Ins. Group, 148 Ga. App. 156 (251 SE2d 48) (1978).
I do not conclude whether or not the “natural inference” may be available in other types of mental or emotional stress cases. For example, this case does not deal with a claim that a sudden, brief excitement or stimulus caused or precipitated an attack of some nature *33which immediately followed the event. See Travelers Ins. Co. v. Neal, 124 Ga. App. 750 (186 SE2d 346) (1971).