Error is claimed in this case by plaintiff-appellant. She is the widow of a decedent who died of a “heart attack” while at work in the manufacturing plant of the defendant-appellee. The claim *228was filed with the workmen’s compensation department. The hearing referee awarded the death benefit provided in the act.1 The workmen’s compensation appeal board reversed and disallowed the award. The case is before ns on our leave granted.
The scope of our review has been legislatively limited. The findings of fact of the appeal board are conclusive in the absence of fraud.2
The immediate question is the entitlement to benefits.
The principle of review involved is whether where the administrative body makes a determination of fact, binding upon us, but includes in its opinion an erroneous statement of law, we should affirm on the basis of the facts found, or reverse and remand for reconsideration by reason of the incorrect legal test stated.
The problem is difficult. We are called upon, in reality, to define the thought-decisional process of the board. This is to ask whether the finding of fact would have been the same irrespective of the misconception of law that is reflected in the majority holding of the divided board.
There is no factual conflict as to what plaintiff was doing at the time of the fatal cardiac. He was drilling 3/4" holes in sheet metal, with a 15- to 20-pound drill requiring 50 pounds of pressure on decedent’s part to push the drill through the metal. He was observed by a witness doing this in apparent total norrhality approximately 45 seconds before his collapse. The same witness, a testing development engineer, turned and walked back to the deceased’s station and saw him on the floor. The drill was still held in both his hands. The bit was wound up in his trouser leg. The half-horse power motor power*229ing the drill was stalled. He appeared to be vomiting bnt seemed to the witness conscious, though unable to talk. He was taken to the plant medical facility and pronounced dead on arrival. The deceased was 49 years old, 5 feet, 6 inches tall, and weighed 176 pounds. He had worked for defendant in all some 13 years. He was in apparent good health at all times prior to the fatality, including that morning when he left for work. After the foregoing factual background was made of record, the medical witnesses were examined on the basic question of the relationship of the work plaintiff was doing to the cardiac failure. Two physicians testified for defendant, one for plaintiff. Their testimony wras in diametric conflict.
The first defense doctor testified in response to a properly prepared hypothetical question:
“A. My opinion is there would be no relationship between the work and the heart attack.”
The second defense expert to essentially the same question replied: “No, I don’t think so.” Each doctor detailed the reasons for his answer. They pointed out that the effort required of decedent to man the drill was no greater than the demands of ordinary living, walking, and similar normal activity. Both alluded to the fact that many fatal cardiacs occur while the subject is in bed asleep; that the effort here involved was totally unrelated to the death.
Plaintiff’s medical witness countered:
“I think there is no question that a heart attack can occur at rest as well as during exertion, whether it be moderate or considerable. * * * In my opinion, * * * any exertion which is coincidental with a heart attack * * * can be an aggravating factor in relation to the ultimate course of the patient.” (Emphasis this Court’s.)
*230To the specific inquiry:
“In your opinion, could this activity that I described this man doing at the time have contributed to the man’s death?
“A. It could have.”
Certainly at this point a classic example of pure fact question was framed, and to the workmen’s compensation appeal board the legislature entrusted sole determination. The board answered the question in the negative, in the following language:
“Can we, with reasonableness, say that Mr. Zaremba’s work on March 16, 1959, was of such a character as to warrant the deduction that it precipitated his heart attachf We think not. This conclusion was not reached precipitously but only after much consideration and upon review of Supreme Court cases and appeal board decisions dealing with the same problem. The connecting causative thread between work and disability or death has been the strenuous character of the work being done. The mere fact of working at the time of disablement has never been deemed sufficient, in and of itself to satisfy the requirement of provable relationship. It is not an ‘ipse dixit.’
“We are, therefore, constrained to hold that the death of Mr. Zaremba was not causally related to his work, and, we so find.” (Emphasis this Court’s.)
We believe the legal test that should control here was stated in Sheppard v. Michigan National Bank, 348 Mich 577. In that case, out of 7 Justices participating, 6 separate opinions were written. Of the 7 Justices, 5 wrote to affirm the award. Two wrote to vacate it. Under the facts there presented, then, a majority of this Court held minimally that plaintiff suffered a disabling and compensable injury. Former Mr. Justice Talbot Smith, in his opinion,3 *231wrote for the adoption of the rale enunciated by the Arkansas supreme court in Bryant Stave & Heading Co. v. White, 227 Ark 147 (296 SW 2d 436):
“ ‘Notwithstanding anything we may have said in prior cases, we hold that an accidental injury arises out of the employment when the required exertion producing the injury is too great for the person undertaking the work, ivhatever the degree of exertion or the condition of his health, provided the exertion is either the sole or a contributing cause of the injury. In short, that an injury is accidental when either the cause or result is unexpected or accidental, although the work being done is usual or ordinary.’” (Emphasis this Court’s.)
It is difficult to tell from the 5 separate opinions exactly what the case stood for then and what it stands for now. If it does not stand for the foregoing quoted rule, I think it ought to. The rule seems to me to reflect the legislative intent of our statute. I write for its adoption or affirmation here.
Since the foregoing rule is the test of causality in an injury case, we must a fortiori apply it to a case where the death ensues.
Thus it is abidingly clear that under this rule the appeal board was perfectly free to accept as more credible the testimony of the defendant’s doctors. It was perfectly free to find that the exertion required of decedent in his work at the time was totally unrelated to the cardiac catastrophe.
It is equally clear that the board was not free to conclude legally that:
“The connecting causative thread between work and disability or death has been the strenuous character of the work being done.” (Emphasis supplied.)
How to disengage this erroneous concept from the board’s fact-finding process, legislatively removed from judicial review, eludes us. We would *232perforce have to ascertain whether the majority in its reasoning to a factual conclusion posited first the “strenuous character” concept test to the non-medical fact background of the type of work being performed and then considered the opinion evidence of causality or lack thereof, in light of that first finding. This we cannot undertake. The reasoning process of the board is relayed to us'by words. The words they used clearly, we think, impart a consideration by the majority hoard of a fallacious test. At what precise point this test was applied, we cannot determine.
For this reason we vacate the order of the board and remand the cause for reconsideration upon the record already made, to determine whether under the test we here established or affirmed, stripped of the strenuous character test, decedent’s work, whatever its nature, was causally related to the fatality. Remanded for further proceedings in accordance herewith. Costs to appellant.
T. M. Kavanagh, C. J., and Souris, Smith, and Adams, JJ., concurred with O’Hara, J.CLS 1961, § 412.5 (Stat Ann 1960 Rev § 17.155), sinee amended by PA 1965, No 44 (Stat Ann 1965 Cum Supp § 17.155).
CL 1948, § 413.12 (Stat Ann 1960 Rev § 17.186).
Sheppard, supra, pp 582, 583.