This appeal brings before us for the second time the celebrated case of Helen Stoleson and her "dynamite heart." See, e.g., Time, July 12, 1971, at 41. The issues on appeal this time concern the causal relationship, if any, between the defendant's now conceded negligence and the symptoms that Mrs. Stoleson has continued to experience long after the elimination of their organic cause.

Mrs. Stoleson, now 64 years old, began working in a federal munitions plant in Wisconsin in 1967 as an employee of the contractor operating the plant. Within a few months she began experiencing the characteristic chest pains of coronary artery disease--but, oddly, only on weekends. One weekend in February 1968 the chest pains were so severe that she was hospitalized. She was diagnosed as having suffered either an actual heart attack (myocardial infarction) or an episode of coronary insufficiency (meaning that the coronary arteries were not supplying the heart with an adequate supply of blood). She returned to work shortly after this incident but continued having weekend chest pains with increasing frequency till she left the plant in 1971.

Her work at the plant required her to handle nitroglycerin and she became convinced that this was causing her heart problem. But the doctors she consulted rejected her theory until she came under the care of a Dr. Lange in 1971. He was convinced by her experience and that of several of her coworkers, who had similar symptoms, that excessive exposure to nitroglycerin had caused their coronary arteries to expand--much as nitroglycerin tablets given for the treatment of coronary artery disease do--and that the sudden withdrawal of nitroglycerin on the weekends had caused the arteries to contract violently. See Lange, et al., Nonatheromatous Ischemic Heart Disease Following Withdrawal from Chronic Industrial Nitroglycerin Exposure, 46 Circulation 666 (1972).

Mrs. Stoleson brought suit under the Federal Tort Claims Act, 28 U.S.C. Secs. 1346(b), 2671 et seq., in 1974, alleging that the government had been negligent in failing to protect the workers at the plant from excessive exposure to nitroglycerin. The district judge dismissed the suit at the close of the plaintiff's evidence, on the basis of the statute of limitations, but this court reversed, Stoleson v. United States, 629 F.2d 1265 (7th Cir.1980), and the case was retried in 1981. At the conclusion of the retrial the district judge found that the government had been negligent and that its negligence had caused Mrs. Stoleson's heart disease, and he awarded her $53,000 in damages. But he declined to award any damages for her psychosomatic illness after she left the plant, and she appeals.

Although Mrs. Stoleson's heart disease should have abated completely soon after she left the plant, her health at the time of the second trial was poor. She had chest pains, though not so acute as when she was working at the plant and not more frequent on weekends than at other times; she was often dizzy and short of breath, and easily fatigued; she had bouts of high blood pressure, and coughing spells leading to vomiting; she was extremely obese, having gained 100 pounds in 10 years; and she was unshakably convinced that her health was ruined and that she could no longer work full time. These complaints have no organic basis, unless it is her obesity. Her heart disease from nitroglycerin exposure did no medically significant lasting damage. Dr. Goldbloom, the psychiatrist who testified for her, diagnosed her condition as "hypochondriacal neurosis"--what is more commonly referred to just as hypochondria--that had been induced by her heart disease at the plant, particularly the possible heart attack in February 1968, and that had been aggravated both by Dr. Lange's having incorrectly advised her that she had serious, permanent heart damage and by this lawsuit. (Lange did not testify; he died before the first trial.) Dr. Roberts, the psychiatrist who testified for the government, testified that Mrs. Stoleson was indeed a hypochondriac but had probably been one all her adult life.

The district judge found that "the matter of the hypochondriacal neurosis, the presence of which today is agreed upon by both experts, is difficult to resolve. Dr. Roberts seemed to me to be on shaky ground when, based upon his 1981 observations and the other information available to him, he undertook to testify that this neurosis existed prior to February, 1968. On the other hand, while intending no disrespect, Dr. Goldbloom seemed to me to be scrambling when he undertook to elevate the 1968 heart episode itself to the level of a substantial factor in causing the neurosis. I do not believe plaintiff met the burden of proof in this latter respect." The judge found that if Mrs. Stoleson had proved a causal linkage between the defendant's negligence and her neurotic symptoms, she would have been entitled to additional damages for lost earnings and for pain and suffering (physical and emotional) of $238,000. A predicate for this finding was the judge's view that these symptoms had begun no earlier than November 1975, when the first trial had ended. Mrs. Stoleson disagrees with this predicate but does not challenge the adequacy of the $238,000 damage figure.

The district judge's finding on causation presents an interpretive problem. It can be read to mean that he thought the important thing was whether Mrs. Stoleson's possible heart attack in February 1968, which was due to the government's negligence in failing to protect her from excessive exposure to nitroglycerin, had caused her hypochondria, and that if it had not she could not recover damages for her hypochondriacal illness. So read, the finding would be inconsistent with the "thin skull" or "eggshell skull" or "you take your victim as you find him" rule of the common law. The substantive law of Wisconsin is conceded to govern this case, see 28 U.S.C. Sec. 1346(b); and, by an odd coincidence, what has come to be the leading case announcing the eggshell skull rule is a Wisconsin case, Vosburg v. Putney, 80 Wis. 523, 50 N.W. 403 (1891), though it is not the earliest eggshell skull case even in Wisconsin, see Stewart v. City of Ripon, 38 Wis. 584, 590-91 (1875). In Vosburg, one school boy kicked another in the shin, in circumstances that made the kicking a battery. The kick would not have seriously injured a normal person, but the victim had an infection in his tibia and the kick aggravated the infection, causing serious injury. The court held the defendant liable for the entire damages. Although we cannot find any modern eggshell skull cases from Wisconsin, the rule is so well established in tort law (see, e.g., Prosser, Handbook of the Law of Torts 261 (4th ed. 1971)) that the government would have a heavy burden of persuading us that Wisconsin has abandoned it, and as a matter of fact has made no effort to persuade us of this.

It would therefore make no difference to the extent of the government's liability that a normal person in Mrs. Stoleson's situation would have recovered her health completely once she found out that her heart disease had been caused by an environmental factor that had been eliminated before it could do any permanent damage. That Mrs. Stoleson's vulnerability was psychological, rather than, as in Vosburg, physical, is irrelevant. Thomas v. United States, 327 F.2d 379, 381 (7th Cir.1964); Mizell v. State, Through Louisiana Dept. of Highways, 398 So. 2d 1136, 1143 (La.App.1980). And though a distinct principle of Wisconsin law bars recovery of damages where the only consequence of the defendant's negligence is emotional distress, McMahon v. Bergeson, 9 Wis. 2d 256, 272, 101 N.W.2d 63, 71 (1960); Ver Hagen v. Gibbons, 47 Wis. 2d 220, 227, 177 N.W.2d 83, 87 (1970), with an immaterial exception discussed in La Fleur v. Mosher, 109 Wis. 2d 112, 325 N.W.2d 314 (1982), this principle is inapplicable here, both because Mrs. Stoleson's hypochondriacal illness is alleged to have been the result of the physical injury she sustained from the nitroglycerin exposure and because the hypochondriacal illness involved physical symptoms.

Even if the February 1968 episode did not trigger Mrs. Stoleson's hypochondriacal symptoms, the government might still be liable if the symptoms were triggered by Dr. Lange's having alarmed her, whether or not he committed professional malpractice in exaggerating the extent of her organic impairment. A principle distinct from that of Vosburg but equally the law in Wisconsin makes a tortfeasor liable for aggravation of the injury he inflicted, even aggravation brought about by the treatment--even the negligent treatment--of the injury by a third party. See Heims v. Hanke, 5 Wis. 2d 465, 471, 93 N.W.2d 455, 459 (1958); Butzow v. Wausau Memorial Hospital, 51 Wis. 2d 281, 187 N.W.2d 349 (1971). If a pedestrian who has been run down by a car is taken to a hospital and because of the hospital's negligence incurs greater medical expenses or suffers more pain and suffering than he would have if the hospital had not been negligent, he can collect his incremental as well as his original damages from the person who ran him down, since they would have been avoided if that person had used due care. So here, if the government had been careful Mrs. Stoleson would have had no occasion to consult Dr. Lange and might therefore not have become a hypochondriac.

If we were persuaded that the district judge had misapplied any of these principles the logical course would be to remand the case for new findings, but we think an alternative reading of his findings on causation is not only possible but more plausible in light of the evidence: that he was unpersuaded that Mrs. Stoleson had established causation as a matter not of law but of fact. On this reading, when he referred to Mrs. Stoleson's "hypochondriacal neurosis" he meant her symptoms, not the underlying psychological condition on which a traumatic event might act to produce symptoms of ill health. This usage would not be surprising. Hypochondria is usually defined in terms of its symptoms rather than its underlying psychological structure, which anyway is not well understood. "Hypochondriasis is a term used to refer to a psychoneurotic pattern of overconcern and focusing of interest on sickness or bodily symptoms. Hypochondriacal individuals are preoccupied with the functioning of their bodies and tend to interpret even minimal alterations or disturbances as ominous portents of deadly diseases or other impending disaster. Their attention may be focused at various times on breathing, bowel movements, the color of their tongues, or a host of minor aches, pains, and other sensations which in fact are within the range of normal. The essential element is a deep conviction that illness exists, which no amount of reassurance can dispel." 4 Gray, Attorneys' Textbook of Medicine p 178.14 at p. 178-9 (3d ed. 1982). "Hypochondriasis, or hypochondria, is an exaggerated concern for one's health that is not based on any physical illness, although the patient feels ill." The Harvard Guide to Modern Psychiatry 35 (Nicholi ed. 1978). "[I]t is a chronic disorder characterized by a persistent preoccupation with the functions of one's body and intractable fears that one is suffering from physical illness." Id. at 195. "Little is known about the cause, nature, or effective treatment of hypochondria." See id. at 195-96.

Mrs. Stoleson's counsel acknowledged at oral argument that it would be appropriate to require a plaintiff to prove by clear and convincing evidence that the defendant's negligence had caused hypochondriacal symptoms. We treat this acknowledgment not as a binding concession upon which to base decision of this appeal but as a recognition of the dangers of allowing proof of hypochondria to magnify the damages in personal injury suits. Since the physical symptoms of hypochondria have by definition no organic basis, and since so "little is known about the cause, nature, or effective treatment of hypochondria," the condition is impossible to diagnose with confidence. There is no way of verifying Mrs. Stoleson's claim that she has chest pains, or of confidently attributing them to hypochondria rather than to an undiagnosed physical condition. The claim invites the trier of fact to speculate. Moreover, if a person has a tendency to transform mental distress into physical symptoms, awarding damages for hypochondria too freely might aggravate those symptoms and thus make litigation itself a source of illness. This danger is underscored by Dr. Goldbloom's testimony ascribing Mrs. Stoleson's hypochondria in part to "the difficulties that she was encountering pursuing a lawsuit [this lawsuit] against the United States," remarking her "involvement of too much energy in these legal proceedings," and acknowledging that "winning the lawsuit ... in the sense of feeling that her position has been vindicated by a higher authority" could lead to a big improvement in her health. Dr. Roberts thought that both Dr. Lange and Mrs. Stoleson's lawyer had aggravated her physical symptoms by encouraging her to sue the government; Roberts accused Lange of "encouraging her belief in her disability."

Without going so far as to hold that liability for hypochondria must be proved by clear and convincing evidence, we advise our district judges to approach such claims with the healthy skepticism necessary to prevent excessive and unfounded damage awards, bearing in mind that there is not much difficulty in finding a medical expert witness to testify to virtually any theory of medical causation short of the fantastic. It is only a partial reply that there is no right to jury trial in Federal Tort Claims Act suits unless individual defendants are joined and demand it. 9 Wright & Miller, Federal Practice and Procedure Sec. 2337 at p. 130 (1971).

Evaluated in light of these general considerations, the finding by the able and experienced district judge that Mrs. Stoleson failed to prove a causal linkage between the government's negligence and her present ill health must be affirmed. The testimony of both experts was speculative, neither having examined her before 1980--nine years after she left the munitions plant. Dr. Goldbloom's testimony was also inconsistent. He began by saying, "I think that the whole thing started with a heart attack that she suffered in 1968," but then he said that the onset of her hypochondriacal symptoms "probably began sometime after Dr. Lange's death" in 1972. Dr. Goldbloom's uncertainty about the date of the onset allowed the district court to conclude that Mrs. Stoleson had not proved that it preceded the end of the first trial in November 1975. Hypochondria had not been so much as mentioned at the first trial, though the trial had taken place four years after she left the plant and she had put in her entire case on damages as well as liability before the suit was dismissed.

With the date of onset pushed forward to November 1975 or later it became a matter of conjecture whether Mrs. Stoleson's hypochondriacal illness really was the delayed consequence of a "dynamite heart" problem that ended in 1971 or whether it was the result of other stress factors, including the litigation itself, which Dr. Goldbloom testified had contributed to her continuing ill health. It would be strange if stress induced by litigation could be attributed in law to the tortfeasor. An alleged tortfeasor should have the right to defend himself in court without thereby multiplying his damages more than five-fold--the damage multiple if Mrs. Stoleson prevails on this appeal.

There are other grounds for doubting whether the causal connection between the government's negligence and Mrs. Stoleson's present condition was proved. Until the second trial Dr. Goldbloom was diagnosing that condition not as hypochondria but as depression and paranoia. Mrs. Stoleson has dropped the contention that her depression and paranoia were caused or aggravated by the government's negligence but it is admitted that she does suffer from depression, paranoia, and delusions (both doctors mentioned her belief that the government had caused or at least hastened Dr. Lange's death from cancer to punish him for exposing the government's negligence), and all of these could be factors in her bad health. Furthermore, Dr. Roberts testified without contradiction that the only one of Mrs. Stoleson's symptoms that is unequivocally a cardiac symptom--a pain in the left side of her chest that radiates down her left arm--she herself attributes not to heart disease but to some undefined, and apparently nonexistent, lung ailment. Dr. Roberts thought many of her symptoms might be due to her obesity, which he refused to ascribe to her hypochondria, and to her poor mental health, which had worsened after and he thought because of her mother's death in 1975. He also testified that hypochondria is a reaction to stress and that he suspected that Mrs. Stoleson had reacted to stress hypochondriacally throughout her adult life. Because Dr. Roberts had no evidence that Mrs. Stoleson had thought herself in ill health before 1967 the district court thought him on shaky ground in testifying that her hypochondria had preceded her exposure to nitroglycerin. But the burden of proof was on her, and Dr. Roberts' testimony, shaky as it may have been in some respects, went far to rebut an inference that her hypochondria had been caused by nitroglycerin.

There is another reason for concluding that Mrs. Stoleson failed to carry her burden of proof. A companion principle to the eggshell skull rule, discussed in our recent decision in Abernathy v. Superior Hardwoods, Inc., 704 F.2d 963 at 973 (7th Cir.1983), is that in calculating damages in an eggshell skull case the trier of fact must make an adjustment for the possibility that the preexisting condition would have resulted in harm to the plaintiff even if there had been no tort. The necessity for such an adjustment was argued in Vosburg although rejected on the facts. But in Abernathy, where the victim of a tortious back injury already had a back disease that might eventually have produced symptoms similar to those produced by the tortious injury, we held that this possibility had to be considered in calculating the victim's damages from the tort. We were interpreting Indiana law in that case but Wisconsin law is to the same effect. See Helleckson v. Loiselle, 37 Wis. 2d 423, 430, 155 N.W.2d 45, 49 (1967). If we assume, as the judge and both medical witnesses did, that Mrs. Stoleson has long been prone to exaggerate her health problems, it is likely that sooner or later some symptom unrelated to the defendant's misconduct would have triggered the neurosis and led to psychosomatic symptoms similar to those she is suffering from. Few people go through middle age without some health problems and a good deal of other stress. If Mrs. Stoleson was so vulnerable that a heart ailment from which she recovered completely, in circumstances that gave no rational basis for thinking it would recur, nevertheless caused the permanent and complete disability of which she complains, the chances are that some lesser trauma, highly likely for a woman of her age, would have caused some lesser disability; and the expected costs of this nontortious injury would have to be subtracted from the damages that she claims. She made no attempt to make this calculation. We do not question the district court's finding that the damages from her neurotic condition amount to $238,000, but not all of this amount can be attributed to the defendant's misconduct, and it was her burden to show what part could be.

Mrs. Stoleson's claim for damages from hypochondria may also be barred on grounds of remoteness of damage. Even when it is reasonably certain that a tort made a particular adverse consequence more likely and that the consequence in fact came to pass--even if, in short, there is no question of cause and effect--the tortfeasor may be excused from liability under Wisconsin law because "the claim of damages ... is so remote and so out of proportion to the culpability of the tortfeasor that, as a matter of public policy, [the court] conclude[s] that the defendants are not to be held liable for this element of damages." Howard v. Mt. Sinai Hospital, Inc., 63 Wis. 2d 515, 519, 217 N.W.2d 383, 385, reh. denied, 63 Wis. 2d 515, 523a, 219 N.W.2d 576 (1974). In Howard an intern negligently allowed a catheter that he was inserting into the plaintiff's shoulder to break off, and two pieces remained in the plaintiff's body. The plaintiff claimed that as a result of the incident she developed an irrational fear of getting cancer. Though it was undisputed that the intern had been negligent, that the plaintiff had developed the fear of which she was complaining, and that the intern's negligence was a substantial factor causing the fear, the Wisconsin Supreme Court held that she could not recover damages for the fear. The decision cannot be explained on the basis of the principle noted earlier that in Wisconsin the negligent infliction of purely emotional injuries is not actionable; the plaintiff in Howard was suing for the physical as well as emotional consequences of the breaking off of the catheter in her arm. The decision seems to stand for an independent principle limiting the scope of liability; and since its facts are not unlike those of the present case we are surprised that the government did not cite Howard to us. Although there is no impropriety in relying on a ground not argued by the parties to affirm, the vagueness of the test employed by the Wisconsin Supreme Court in Howard makes it difficult to predict how that court would decide the present case. Howard merely gives us some additional confidence that we are correct in concluding that the judgment appealed from must be

AFFIRMED.