WISDOM, Circuit Judge.

This case involves insurance coverage under a double indemnity clause when the insured is a man of advanced years and has certain diseases usually associated with old age. The question before us is whether there is substantial evidence to justify the jury finding that the insured’s death resulted from an *769 accidental injury, directly and independently of all other causes, and did not result directly or indirectly from disease.

William B. McGehee of Montgomery, Alabama, was old and blind. January 5, 1957, going from his bed to a chair with the help of his wife, he fell and skinned a knee. The abrasion would not heal. McGehee was hospitalized. Acute cellulitis (infection of the skin and tissue) developed. 1 Six days after the fall his attending physician, Dr. William Cawthon, said that the cellulitis was the worst case he had even seen. The infection sent McGehee’s blood count to more than twice the normal count. He ran a high fever. He had trouble breathing. His legs swelled. The abrasion became inflamed and red streaks ran down his leg from the knee. His heart weakened. Unable to cope with all the complications that were taxing his resistance, he died of heart failure sixteen days after the fall. He was seventy-seven years old.

McGehee had carried two noncancella-ble insurance policies with the New York Life since 1927. The company admitted liability for the face amount of the policies, but denied liability for double indemnity. The double indemnity clause in each policy reads, in part:

“The Double Indemnity * * * shall be payable upon receipt of due proof that the death of the Insured resulted directly and independently of all other causes from bodily injury effected solely through external, violent and accidental means and occurred within ninety days after such injury.

“Double Indemnity shall not be payable if the insured’s death resulted * * *, directly or indirectly, from infirmity of mind or body, from illness or disease, or from any bacterial infection other than bacterial infection occurring in consequence of accidental and external bodily injury.”

The New York Life takes the position that the insured's death did not result from his fall, “directly and independently of all other causes”; that at the time of the accident McGehee had latent diabetes, arteriosclerosis, heart disease, emphysema, and arthritis, any or all of which contributed to or caused his death.

An autopsy report 2 stated that death was “attributed to heart failure with the heart disease being on an arteriosclerosis and probably hypertensive basis”. At one time McGehee had been diabetic, but he had not taken insulin for four years prior to the fall; he was given insulin in his last illness. In filling out the death certificate Dr. Cawthon gave arteriosclerotic heart disease and coronary artery disease as the “Disease or Conditions Directly Leading to Death and Antecedent Causes”. However, under the heading “Accident, Suicide, Homicide” on the certificate, Dr. Caw- *770 thon wrote: “Fell, abraiding right knee.” In explanation he testified that he equated the accident with suicide and homicide. 3 Dr. Cawthon stated: Me-Gehee “would not have died had he not had the accident and the infection. * * He had things a man 77 years old has * * * there were things wrong with him, as there would be in anybody, but the essential thing without which death would not have occurred is that, the infection”.

The ease was tried on the sole issue of whether McGehee’s death resulted directly and independently of all other causes from bodily injury effected through external, violent, and accidental means. The jury returned a verdict for $30,000 for the Executrix of the Estate of William B. McGehee. The defendant insurance company specifies as error the failure of the trial court to direct a verdict for the defendant, failure of the court to give judgment for the defendant notwithstanding the verdict, and failure tp order a new trial. We affirm the judgment of the district court.

Clauses similar to the double indemnity clauses in McGehee’s policies have been considered by the courts on many occasions. The authorities are conflicting and their dicta confusing. One of the cases most frequently cited is Silverstein v. Metropolitan Life Ins. Co., 1930, 254 N.Y. 81, 171 N.E. 914, 915, in which Mr. Justice Cardozo made the observation, pertinent in this case, “[A] policy of insurance is not accepted with the thought that its coverage is to be restricted to an Apolio or a Hercules”.

Alabama law controls our decision. The leading case is probably First National Bank of Birmingham v. Equitable Life Assurance Society, 1932, 225 Ala. 586, 144 So. 451, 453, in which the court reviewed the authorities, noting that the policy in question had a special clause providing that there would be no coverage where death results “directly or indirectly from disease or bodily infirmity”; “some effect must be given to such provisions”. The Alabama court concluded that the question is one of proximate cause. 4 Proximate cause “does not [necessarily] mean the cause or condition which is nearest in time or space to the result. * * * The law does not consider the cause of causes beyond seeking the efficient predominant cause”. For an insurance company to escape liability for double indemnity, a disease must be the “efficient, procuring, cause” of the injury or death. And disease does not mean “mere feebleness, nor predisposition to recurrence of former disease, nor every infirmity which may aggravate the effects of an accidental injury”.

Later decisions have not modified the principles expressed in Equitable *771 Life Assurance. The cases turn on the question of causation, a factual issue that is for the jury to decide.

In Emergency Aid Ins. Co. v. Connell, 1952, 258 Ala. 521, 63 So.2d 603, 605, an insured broke his leg in 1935. It did not heal properly and became infected. In 1948 he injured the same leg, and it had to be amputated. The Court said: “We are not willing to say as a matter of law that the condition of his leg was an efficient cause of the injury or disability which he suffered, although it may have aggravated it. It had no relation whatsoever to bringing about the accident itself. The jury could find that it was only a condition upon which the accident in June 1948 operated, although the serious results of that accident were aggravated by it”.

In John Hancock Mut. Life Ins. Co. v. McCreary, 1954, 37 Ala.App. 493, 70 So.2d 817, 820 an insured, sixty-three years old, had Parkinson’s disease. The double indemnity clause was almost the same as the one in McGehee’s policies. McCreary suffered bruises and abrasions from an automobile accident. The death certificate showed pulmonary emboli as the direct cause of McCreary’s death, a thrombosis as an antecedent cause, and listed Parkinson’s syndrome as a significant condition. There was medical testimony that Parkinson’s disease contributed to McCreary’s death in that a well man would not have died of the accident, but that the pulmonary embolism would have killed him and that the embolism would not have developed without some trigger mechanism (trauma or infection) causing it to develop. The Court held for the insured:

“When an insured suffers an accident independent of and unconnected with a disease from which he is suffering, it is ordinarily a question for the jury to determine, where the evidence is conflicting, whether the accident was the sole proximate cause of the injury or death, even though the effects thereof may have been aggravated by the disease. * * * [T]he jury could reasonably infer, and were amply justified in so doing, that such bruises and abrasions resulted from the accident, that this accident was the proximate cause of the femoral thrombus with the resulting emboli which directly caused the insured’s death.”

See also Emergency Aid Insurance Co. v. Dobbs, 1955, 263 Ala. 594, 83 So.2d 335, approving McCreary.

Cases denying recovery are not inconsistent with McCreary and Connell. In Equitable Life Assurance the court held, as a matter of law, that an old injury to the temporal lobes of the insured’s brain producing abscess of the brain and adhesions was an efficient cause of the death. Similarly, in Adkins v. Metropolitan Life Ins. Co., 235 Ala. 417, 179 So. 382, the court also held, as a matter of law, that “Addison’s disease”, a fatal disease, was the cause of the insured’s death, and that the injury was not caused by accidental means. See also Provident Life & Accident Ins. Co. v. Nelson, 38 Ala.App. 372, 84 So.2d 130; Metropolitan Life Ins Co. v. Halsey, 230 Ala. 193, 160 So. 248; Prudential Insurance Co. v. Calvin, 227 Ala. 146, 148 So. 837.

This Court has dealt with the problem before. In Ryan v. Continental Casualty Co., 5 Cir., 1931, 47 F.2d 472, a pre-Erie case, the insured fell while loading ice on a truck. Within an hour of the accident he died as the result of a ruptured aorta. A fall, no matter how violent, will not rupture an aorta. The Court held for the insurer. On the facts it was clear that the fall was not the proximate cause of the insured’s death. In Davis v. Jefferson Standard Life Insurance Co., 5 Cir., 1934, 73 F.2d 330, 96 A.L.R. 599, also before Erie, the insured died after an anaesthetic was administered in the course of an operation for an old hernia. This Court denied recovery of double indemnity on the ground that the administration of the anaesthetic was not accidental. The insurance was not against unintended death but death from injury by acci *772 dental means. The insured’s widow claimed also that her husband’s death was caused by his being shaken up in an automobile accident three days before the operation. He had no bruises and had not complained of any heart trouble. This Court held that it was unreasonable to conclude that the dilation of the insured’s heart on the operating table was connected with the alleged automobile accident. In Mutual Life Insurance Co. of New York v. Loeb, 5 Cir., 1939, 107 F.2d 7 the insured cut her hand on a kitchen cabinet. An infection appeared. She was taken to the hospital, where she died. Every physician who testified stated that her death was from a combination of causes: diabetes, chronic nephritis with uremia, staphylococcus in the blood stream. At the time of the accident she was receiving total disability benefits for chronic nephritis and arteriosclerosis and, three months before the accident, was reported as suffering from advanced cardio vascular renal disease and chronic diabetes. The Court found that a verdict should have been directed for the insurer.

New York Life Ins. Co. v. Schlatter, 5 Cir., 1953, 203 F.2d 184, 187, was a Mississippi case but it involved a double indemnity clause reading exactly the same as the clause in McGehee’s policies. A woman sixty-three years old fell and broke her hip. She was taken to the hospital, operated on ten days later, and died within twelve hours of the operation. It was undisputed that at the time of the accident Mrs. Schlatter was suffering from high blood pressure, heart trouble, and various other ailments. Her physician testified that her heart condition was a contributing cause of her death. When asked whether she could have died from the heart condition irrespective of the accident, he answered: “She could have died or could have lived for years * * * [T]he broken hip is what brought the whole thing on— started the momentum.” Judge Rives, speaking for the Court, held: “[T]he question is largely one of proximate cause. So construing the policy, it seems to us that the evidence made a case for the jury.” The Court concluded that the rule in United States Fidelity & Guaranty Co. v. Hood, 124 Miss. 548, 87 So. 115, 15 A.L.R. 605 was controlling. 5 In a later case Judge Rives observed that: “In Alabama the phrase ‘solely through external, violent and accidental means’ has been construed ‘to mean that the accident shall be the proximate cause of death and not exclusive of other conditions, means, or circumstances’ ”. Gay v. Pacific Mutual Life Ins. Co., 5 Cir., 1956, 237 F.2d 448, 450. See also New York Life Insurance Co. v. Johnston, 5 Cir., 1958, 256 F.2d 115.

A review of the Alabama cases and the decisions of this Court shows that considerable latitude must be allowed the jury in determining the question of causation. An insurer cannot escape liability simply by showing that some disease may have contributed to some extent to the insured’s death. Most elderly people have some ailment, some latent disease. When an old person is injured, almost invariably an ailment is aggravated, a dormant disease activated, and the effects of the injury intensified because of general frailty and lack of resistance to illness. Progressive weakness and increasing complications follow any time an old person is put to bed for any length of time. In every case therefore it is difficult to separate the effects of the accident from the effects of disease. The insurer would have us hold that there can be no recovery unless Death is present at the scene of the accident, or openly waits in continuous attendance on an injured insured from the *773 moment of injury to the moment of death. Alabama courts, and this Court, take a more reasonable view. If an insured has an active disease of such a character as to endanger the insured’s life, apart from the accident, such a disease is a contributing cause that will bar recovery. If however an injury starts a chain reaction resulting in death, recovery may be allowed even if one of the links in the chain is old age frailty and some of the links are dormant diseases or physical conditions without which the chain would be broken. “The answer to the question of whether substantial evidence supports the verdict depends, of course, on the details of the evidence in each particular case * * * ”. New York Life Insurance Co. v. Johnston, 5 Cir., 1958, 256 F.2d 115, 117.

William McGehee was blind and seventy-seven years old. But Dr. Scott, who performed the autopsy, described McGehee as “a middle aged man”, and Dr. Cawthon, physician for McGehee’s last fifteen months, found his patient’s physical condition unchanged until the accident. Dr. Cawthon told Mrs. Mc-Gehee that her husband could live ten years. McGehee had not been treated for diabetes, a mild case, for four years. Before the fall, he had no symptons of diabetes while Dr. Cawthon was treating him. The autopsy showed that the “Isles of Langerhans” in the pancreas were of normal size, shape, and position; usually diabetes destroys the Isles of Langerhans. McGehee’s arthritis was only in the hand at the time of the accident, and the autopsy report stated that the insured “arteriosclerosis is moderate”. Arteriosclerosis, hardening of the arteries, is an aging process that starts with birth; Dr. Barnes testified that “a great percentage of people have some, more or less, degree of hardening of the arteries”. As for heart trouble, McGehee had not taken digitalis for five or six months before being hospitalized. The autopsy showed no large myocardial scarring and no evidence of any myocardial infarction. McGehee had no history of angina pectoris. There were no occlusions in the cardio vascular system. There was interstitial fibrosis of the myocardium, but this is not unusual in a man of seventy-seven. Mild emphysema, as reported in autopsy, is common in persons of McGehee’s age group.

The testimony is not all one-sided, but it seems to us on a close reading of the record that the evidence makes a case for the jury.

The judgment is affirmed.

Notes

. Dr. Cawthon described the nature of cellulitis as an inflammation of the skin in response to the presence of an infection. (R. 96) “Tlie presence of this inflammation causes fever, high white blood count, rapid heartbeat, and frequently some prostration, extreme weakness, what wo — what we call toxicity, for lack of a better word, nobody knows exactly what an infection like that, how it does to the body, wliat it does, but it can make you very sick, and it puts a strain on the circulation, heart primarily.”

. The autopsy report listed the following “pathological diagnoses”: 1. Cardiac hypertrophy and dilatation, 2. Arteriosclerosis and arteriolosclerosis, generalized, 3. Hydrothorax, bilateral, 4. As-cites, 5. Chronic passive congestion, mild, 6. Cerebral arteriosclerosis, 7. Chronic Pneumonitis, bilateral. The summary stated: “Death in this case is attributed to heart failure with the heart disease being on an arteriosclerosis and probably hypertensive basis. The generalized arteriolosclerosis supports the latter. The gastric dilatation is considered to be a relatively terminal event. The chronic pneumonitis is focal and no specific etiology is apparent. The right leg was not examined, however, this was felt by the clinician to represent a mild cellulitis. There was no evidence in the autopsy examination to support a septicemia or a pulmonary embolus.”

. “This was the first patient I had ever had occasion to fill a death certificate out on who died as a result of an injury, and I felt that this was the way it was' — that little block there was the place for me to put down the fact that germane to this death and essential to it was the accident, and so I put it down there.”

. The Alabama Court quoted at length from Freeman v. Mercantile Mutual Accident Ass’n, 156 Mass. 351, 30 N.E. 1013, 17 L.R.A. 753, on the meaning of The Maxim. “Causa proxim a non remota speotatur’’ and its application to the factual situation, concluding: “The application of that principle of law to the case at bar is that, if the insured was suffering from a disease which was accelerated and aggravated by the accident so as to be a cause co-operating with it to produce the fatal end, then there can be no recovery. Manifestly recovery is not barred merely because the insured is suffering from disease. One upon a bed of illness may meet death by an explosion or other accidental means. Bohaker v. Travelers’ Ins. Co., 215 Mass. 32, 102 N.E. 342, 46 L.R.A.,N.S., 543. If there is no active disease, but merely a frail general condition, so that powers of resistance are easily overcome, or merely a tendency to disease which is started up and made operative, whereby death results, then there may be recovery even though the accident would not have caused that effect upon a healthy person in a normal state.” First National Bank of Birmingham v. Equitable Life Assurance Society of the United States, 225 Ala. 586, 144 So. 451, 455.

. “We think that, if the accident is the proximate cause of the death and sets in motion or starts a latent or dormant disease, and such disease merely contributes to the death after being so precipitated by the accident, it is not a proximate cause of the death nor a contributing cause 'within the meaning of the terms of the policy.” United States Fidelity & Guaranty Co. v. Hood, 124 Miss. 548, 87 So. 115, 119, 15 A.L.R. 605.