Franklin Delano Roosevelt’s (FDR’s) (1882–1945) 1921 neurological disease revisited;the most likely diagnosis remains Guillain

Abstract

In 2003, we published evidence that the most likely cause of FDR’s 1921 neurological disease was Guillain–Barré syndrome. Afterwards, several historians and neurologists stated in their publications that FDR had paralytic poliomyelitis. However, significant criticism of our article or new support for that diagnosis was not revealed. One critic claimed that FDR’s cerebrospinal fluid indicated poliomyelitis, but we did not find evidence that a lumbar puncture was performed. The diagnosis of FDR’s neurological disease still depends upon documented clinical abnormalities. His age, prolonged symmetric ascending paralysis, transient numbness, protracted dysaesthesia (pain on slight touch), facial paralysis, bladder and bowel dysfunction, and absence of meningismus are typical of Guillain–Barré syndrome and are inconsistent with paralytic poliomyelitis. FDR’s prolonged fever was atypical for both diseases. Finally, permanent paralysis, though commoner in paralytic poliomyelitis, is frequent in Guillain–Barré syndrome. Thus, the clinical findings indicate the most likely diagnosis in FDR’s case remains Guillain–Barré syndrome.

Keywords

Franklin Delano Roosevelt neurological illness paralytic poliomyelitis Guillain–Barré syndrome autoimmune neuropathy Campylobacter cerebrospinal fluid

Introduction

In 2003, we published an article concerning the cause of the paralytic illness that struck Franklin Delano Roosevelt (FDR) in 1921¹ and permanently paralysed his lower extremities (Figure 1). In that investigation,¹ we reviewed the evidence concerning the illness and reconstructed the temporal sequence of the clinical events. In addition to fever and permanent paralysis of the lower extremities, FDR displayed symmetric ascending flaccid paralysis, facial paralysis without involvement of other cranial nerves, transient numbness, prolonged dysaesthesia, significant bowel and bladder dysfunction, lack of meningismus, and recovery of motor function in a descending pattern. We tested whether the illness was most likely the long-held diagnosis of paralytic poliomyelitis or of a different disease. First, his clinical manifestations were documented and compared with those found in non-traumatic causes of flaccid paralysis. Pattern recognition, Bayesian analysis and reconstruction of the pathogenesis of the disease were used to ascertain which disease was most likely. Each method indicated that FDR’s illness was inconsistent with paralytic poliomyelitis^2–6 and was most consistent with Guillain–Barré syndrome (GBS), an autoimmune neuropathy.^3,6–12

Figure 1.

Photograph of Roosevelt taken in 1923 at Warm Springs, Georgia. The muscles of the upper extremities were well developed, whereas those of the lower extremities were symmetrically atrophic, as often occurs in severe cases of Guillain–Barré syndrome but rarely in paralytic poliomyelitis (Courtesy of The Franklin D. Roosevelt Library website; version date 2009).

It would have been helpful if FDR’s cerebrospinal fluid (CSF) had been examined in the first few days of his paralysis because the concentrations of leukocytes and total protein in CSF during the early phase of the illness are quite different in paralytic poliomyelitis and GBS. In poliomyelitis, the numbers of lymphocytes and neutrophils are usually increased significantly, whereas the concentration of total protein is normal (20–45 mg/dL in adults) or slightly elevated.^2–6 By contrast, in GBS the number of leukocytes is usually normal (zero to five mononuclear leukocytes/mm³), whereas the concentration of total protein is increased significantly.^3,6–12 However, the records of FDR’s illness¹ did not produce any evidence that FDR’s CSF was examined.

Criticisms of the 2003 report

Most histories of FDR^13–17 and of poliomyelitis^18–21 and a book on GBS published since our article appeared in 2003¹ still hold that FDR had paralytic poliomyelitis. Ironically, the fairest treatment of our article was in a history of poliomyelitis written for young adults.²¹ Our article was summarized and the possibility was raised that FDR did not have poliomyelitis. By contrast, in Oshinsky’s history concerning poliomyelitis¹⁹ our article was mentioned but summarily dismissed. Oshinsky did not critique the main issues in our article or provide new information to support the diagnosis of poliomyelitis in FDR’s case.¹⁹ Instead, he relied on remarks from a media interview that were taken out of context and raised other unconvincing arguments.¹⁹ He also stated that FDR’s brief immersion in cold water a few days before the onset of his paralytic illness probably had a deleterious effect upon his immune system.¹⁹ Indeed, the opposite may occur following brief, cold water immersion.²² In addition, he felt that FDR’s visit to a Boy Scout Camp predisposed him to poliomyelitis. We deal with that matter later in this paper. In his final analysis, Oshinsky relied upon the diagnosis made by FDR’s physicians in 1921 rather than a consideration of FDR’s symptoms and the current knowledge of neurological diseases that cause flaccid paralysis.

A paradoxical criticism of our article¹ appeared in a book on GBS written for the lay public by two respected neurologists.²³ They rejected the conclusion of our prior publication¹ with the statement ‘most people familiar with GBS feel his [Roosevelt’s] illness was much more typical of polio than GBS’ but failed to identify who ‘most people’ might be or how ‘his [Roosevelt’s] illness was much more typical of poliomyelitis’. Their conclusions were not justified by citations. They also claimed that sensory symptoms occur in paralytic poliomyelitis but that was unsupported by references and runs counter to published descriptions of paralytic poliomyelitis.^2–7 Moreover, other statements contradicted their position concerning FDR’s illness. (1) ‘By the time the average [GBS] patient goes to the doctor, there will be some combination of symmetric weakness, usually in an ascending pattern, and abnormal sensation’. (2) ‘Polio differs from GBS in that the weakness of polio is usually asymmetrical’. Thus, their description concerning the features of GBS indicates that FDR’s symmetric, ascending paralysis was typical of GBS and not of poliomyelitis.

FDR’s 1921 illness was discussed in a 2009 peer-reviewed article.²⁴ Because of FDR’s fever and the expertise of his physicians, Dr Robert W Lovett (1859–1924) and Dr George Draper (1880–1959), the author concluded that the illness was paralytic poliomyelitis.²³ Indeed, in our analysis¹ fever was acknowledged to favour paralytic poliomyelitis^2–7 and so this was already taken into account. However, the fever pattern in FDR’s case was atypical for paralytic poliomyelitis since he did not have a three- to four-day prodromal febrile period before his paralysis began as characteristically is found in paralytic poliomyelitis.^2–7 In addition, the fever in Roosevelt’s case¹ lasted for three to four days longer than the fever typically found during the paralytic phase of poliomyelitis.^2–7

Was a spinal tap performed on FDR?

A proposed rationale for doubting the possibility of GBS and for believing that the disease was poliomyelitis was mentioned in a book concerning FDR’s terminal disease¹⁷ and described in greater detail by one of the authors, a neurologist, in an online commentary.²⁵ He stated that he found in the papers of Dr Robert W Lovett that Lovett requested a spinal tap be performed on FDR. Because of that request, the commentator concluded that a spinal tap was very likely performed and that the CSF findings confirmed the diagnosis of paralytic poliomyelitis.

We also obtained a copy of Dr Lovett’s correspondence dated 24 August 1921.²⁶ Lovett indeed requested a spinal tap in a telegram to the neurosurgeon Dr William Williams Keen (1837–1932) who saw FDR on the third day of the illness.¹ The telegram read:

Leave Boston tonight Tuesday – proceed to Eastport and patients house – desirous to return Wednesday night. Have telegraphed wife of patient – unless already done spinal tapping, cell count, & globular reaction would be useful. This I cannot do personally. Shall arrive at Harvard Club Boston tonight.

Surely if a lumbar puncture had been performed, the procedure and the results of the CSF examinations would have been included in some correspondence and physician reports concerning FDR’s illness? In particular, Dr Lovett’s letter concerning FDR’s paralytic illness to Dr Draper on 12 September 1921²⁷ (a few days before FDR was transferred to Presbyterian Hospital in New York City where Dr Draper, an expert on poliomyelitis, would care for him) did not mention that a lumbar puncture was performed. In addition, the procedure was unlikely for several reasons:

Dr Eben Homer Bennett, the local family physician from Lubec, Maine, would not have performed the lumbar puncture at the start of FDR’s illness because his diagnosis was ‘a heavy cold’.²⁸

Dr William Keen, the neurosurgeon who saw FDR on the third day of the illness, would not have initiated the procedure because his diagnosis was ‘a clot of blood from a sudden congestion – settled in the lower spinal cord.²⁹ Also, it is doubtful that he had the equipment to perform the test since he was vacationing in Maine at the time.

It is improbable that suitable equipment for a lumbar puncture and for CSF examination was available in the small community of Lubec.

It is clear from Lovett’s telegram that he (Lovett) would not perform a lumbar puncture.²⁶

Since the three physicians were from the USA, they would have needed a licence from the Canadian Medical Council to perform the procedure in Canada.

Lovett requested the spinal tap on the 14th day of FDR’s illness and he did not see FDR until the next day. We asked therefore whether CSF findings in paralytic poliomyelitis and GBS are the same at the onset of the paralysis and 15 days later. In fact, the classical distinctions between those two diseases, based upon the concentrations of leukocytes and total protein in the CSF, do not hold after the first several days of paralysis. The evidence is as follows.

In 1966, the CSF findings in paralytic poliomyelitis were re-examined. One hundred and nine cases with type 3 paralytic poliomyelitis were investigated.³⁰ During the first week, 16% had < 10 leukocytes/mm³ in CSF after the onset of the illness. By contrast, during the second week, 38% of them had < 10 leukocytes/mm³ in CSF. In those same patients, during the first week after the onset of the illness, 56% had concentrations of total protein < 45 mg/dl. In the second week, 67% had concentrations of total protein < 45 mg/dl.³⁰ The results indicated that more patients with paralytic poliomyelitis had normal CSF findings a few weeks after the onset of the paralysis than in the first few days of the illness.

The CSF findings in GBS over time were investigated in 24 patients in 1979.³¹ Leukocyte counts in CSF were normal in the first 10 days after the onset of symptoms in all 24 patients. However, 13 of the 24 patients (54%) had an increased number of mononuclear leukocytes in their CSF 10 days to four months after the onset of the paralysis. From the 11th through the 20th day, five of those 13 patients had increased numbers of mononuclear leukocytes in their CSF. Four others developed an increased number of mononuclear leukocytes in their CSF during the next 20 days. In seven of the 13 patients, the numbers of leukocytes in CSF were between 77 to 821/mm³. Between 61 and 120 days, three patients still had increased numbers of leukocytes in the CSF, but the numbers were lower than those found in the previous period. Four months after the onset of the disease, the numbers of leukocytes in the CSF in all 24 patients were normal.³¹ The concentration of albumin in CSF was increased in most patients during the first four months after the onset.³¹ These findings indicate that CSF findings in GBS change over time. Unless the physician is familiar with those changes, the presence of high numbers of mononuclear leukocytes will be confusing in patients with GBS.

Why the diagnosis remains GBS

These findings indicate that, even in the unlikely case that a lumbar puncture was done, it might have been difficult to distinguish between paralytic poliomyelitis and GBS based upon examination of CSF collected two weeks (when Lovett first examined FDR) or four weeks (when FDR was hospitalized in New York City) after the onset of paralysis. Also, since the types of leukocytes probably would not have been identified in Lubec, it would have been impossible to determine whether the pattern of leukocytes was in keeping with poliomyelitis (lymphocytes and neutrophils)^2–6 or with GBS (mononuclear leukocytes).³¹ If many mononuclear leukocytes and a high concentration of protein were found in CSF obtained two to three weeks into a paralytic illness, the findings would be consistent with GBS. On the other hand, if a normal concentration of leukocytes and a normal to slightly increased concentration of total protein were found, then the diagnosis would be more consistent with paralytic poliomyelitis.

One critic was sure that FDR had paralytic poliomyelitis for the following reasons:²⁵

Dr Lovett, the leading expert on poliomyelitis, would have distinguished paralytic poliomyelitis from GBS.

Since FDR became the world’s leading expert on poliomyelitis after Lovett’s death (in 1924), he would have had the same medical expertise and diagnostic prowess.

The timing of FDR’s visit to a Boy Scout camp before the onset of the illness, the muscle tenderness at the onset of the illness and the degree of atrophy lead to the inevitable conclusion that the diagnosis was paralytic poliomyelitis.

Our responses to those comments are as follows:

Dr Lovett’s expertise. Dr Lovett was a leading expert on paralytic poliomyelitis. However, five years before FDR’s paralytic illness, he wrote that Landry’s ascending paralysis³² (later termed GBS) and polyneuritis were among the presenting features of paralytic poliomyelitis.³³ Yet, neither occurs in paralytic poliomyelitis.^2–6 In that respect, the report by Georges Guillain (1876–1967), Jean-Alexandre Barré (1880–1967) and André Strohl (1887–1977) of the CSF and nerve-muscle stimulation abnormalities in the early phase of Landry’s ascending paralysis was published only five years before FDR’s illness.³⁴ Indeed, there is no evidence that Lovett was aware of that publication³⁴ or of the distinctions in the histopathology in paralytic poliomyelitis (damage to motor neurons) and Landry’s ascending paralysis (a peripheral neuritis) reported in 1879.³⁵ Since GBS was never mentioned in any clinical records or correspondence concerning FDR’s case, it is doubtful that Lovett or FDR’s other doctors knew that Landry’s ascending paralysis³² was a separate entity.

FDR’s expertise. FDR had little education in biology and no formal training in medicine. It was in the social not the medical aspects of poliomyelitis that FDR excelled.

Incubation period. The timing of FDR’s visit to a Boy Scout camp on 28 July (about 12 days before the onset of his illness)^13,15 was within the 6–20 day incubation period of poliomyelitis.³⁶ However, although mild or asymptomatic cases could have been missed, there were no reports of poliomyelitis at that camp. Furthermore, the timing was also consistent with a prior infection with Campylobacter jejuni which is the commonest trigger of GBS. Campylobacter jejuni infections are usually recognized because of enteric symptoms.³⁷ Thus, it might seem improbable that such an infection occurred before the onset of FDR’s neurological illness since there were no reports of preceding symptoms. However, in a recent study of the axonal variant of GBS initiated by Campylobacter jejuni,³⁸ diarrhoea before the development of muscle weakness was reported in only 49% of the patients. Thus FDR could have been infected with that enteric pathogen without having gastrointestinal symptoms. It is also possible that a different infection triggered his autoimmune neuropathy.^1,9

Muscle tenderness. FDR did not have muscle tenderness at the onset of his disease. Instead, he experienced severe, protracted dysaesthesia. Dysaesthesia is very common in GBS,^6–12 but rare in paralytic poliomyelitis.^2–7 In that respect, in a recent study protracted pain occurred in 67% of 156 patients with GBS.³⁹

Degree of atrophy. It is unclear whether the commentator meant the degree of atrophy of individual muscles, the number of paralysed (and hence atrophic) muscles or the number of atrophic muscles including those due to paralysis or disuse. Regardless, we did not find any peer-reviewed reports that suggested that patients with paralytic poliomyelitis had more or less atrophy than those with paralysis due to GBS. If he equated atrophy with the frequency of permanent paralysis, then the following is germane. Permanent paralysis is commoner in paralytic poliomyelitis (about 50% of cases)^4,5,7 than in GBS (about 15% of untreated cases).^10,11 However, approximately 35% of patients with severe GBS are permanently paralysed.^40,41 In any case, we took permanent paralysis into account in our previous analysis¹ along with FDR’s other signs and symptoms.

Given these imponderables, sorting out the diagnosis of FDR’s neurological illness eight decades after the fact depended upon an analysis of the clinical features of the illness. However, critics of our article concerning FDR’s 1921 paralytic illness¹ basically ignored the other prominent features of FDR’s illness including his age of onset of the illness (39 years), the prolonged symmetric ascending paralysis, the temporary numbness of the lower extremities, the severe protracted dysaesthesia, the facial paralysis in the absence of other cranial neuropathies, the prolonged inability to void⁴² or defaecate, the absence of meningismus, and the descending pattern of recovery of motor function. All of these clinical features are rare or absent in paralytic poliomyelitis^2–7 but are hallmarks of GBS.^6–12 The symmetry of permanent muscular weakness particularly favours GBS, especially the axonal variant.⁴¹

Finally, our analysis of FDR’s illness was multifaceted.¹ We compared his clinical features with the clinical features of diseases that cause non-traumatic flaccid paralysis. Even without the aid of statistics, pattern recognition indicated the most probable diagnosis was GBS.

The neurological manifestations of FDR’s illness were then used to reconstruct the pathogenesis of his disease. The sensory abnormalities were inconsistent with paralytic poliomyelitis because only motor neurons are invaded by polioviruses, whereas sensory neurons are spared.³⁹ However, sensory and motor nerve involvement is expected in GBS because of an attack by autoantibodies to GM1 gangliosides on those nerves.^43,44 The delay in or failure to repair those peripheral nerves is initiated by those same autoantibodies.⁴⁵

The clinical data were analysed further by the statistical method first proposed by Reverend Thomas Bayes (1702–1761)^46,47 and further developed and popularised by Pierre-Simon Laplace (1749–1827).⁴⁷ Bayesian analysis is a straightforward mathematical formalization of what physicians should do in medical diagnosis. The probability of the occurrence of each disease under consideration and which ones most closely match the symptoms and signs of the patient’s illness are ascertained. We estimated the diagnostic probabilities of flaccid paralysis at the time of FDR’s illness (posterior probabilities) by multiplying prior probabilities of each disease (relative frequency of paralytic poliomyelitis and GBS in FDR’s age group in the early 1920s) by symptom probabilities (likelihood of the patient’s symptoms in paralytic poliomyelitis or GBS). Although we weighed the frequencies of clinical indicators to favour poliomyelitis, the Bayesian analysis confirmed that the most probable diagnosis was GBS.

Other autoimmune neuropathies were also considered in our analysis,¹ including chronic inflammatory demyelinating polyradiculoneuropathy.^48,49 However, the clinical features of GBS, including the symmetric ascending flaccid paralysis, the duration of the progression of the paralysis, and the sensory nerve involvement, matched FDR’s illness far more closely than those found in the other autoimmune neuropathies.

Although we are not absolutely sure of the diagnosis, the clinical evidence strongly indicates the most probable cause of FDR’s neurological disease was GBS.¹ Since our 2003 article,¹ a convincing alternative analysis has not appeared in journal articles or books concerning the cause of FDR’s 1921 neurological illness. Indeed, such an analysis must be based on published data, all aspects of the illness and the pathogenesis of the neurological abnormalities. New studies of that type would be welcomed. Yet, in their absence, many historians and physicians apparently still believe in the diagnosis of paralytic poliomyelitis. A full exploration of this persistent belief in the face of evidence to the contrary is beyond the scope of this paper. However, we offer a few possible explanations.

The first is ‘confirmation bias’, the tendency to favour information that confirms preconceptions or hypotheses regardless of whether the information is true.⁵⁰ The English philosopher and scientist Francis Bacon (1561–1626) observed that ‘the first conclusion colours and brings into conformity with itself all that comes after’.⁵¹ Indeed, it is common to adhere to long-held beliefs and filter out evidence to the contrary. Since expert historians and neurologists repeated many times that FDR had poliomyelitis, it would be difficult for them to accept a different diagnosis.

‘Appeal to authority’ where it is argued that a statement is correct because the person who made the statement is regarded as authoritative, is another possible reason for the persistence of the belief that FDR had poliomyelitis. Expert physicians cared for FDR. However, they were probably unaware that GBS was a separate disease. Furthermore, a consensus concerning the diagnostic criteria for GBS was not established until the latter part of the 20th century.⁵²

Another explanation for the continued insistence that FDR had poliomyelitis is that it makes the history of poliomyelitis more straightforward. For many decades, there has been a simple logical connection between FDR’s illness and FDR’s leadership in the effort to conquer poliomyelitis.

Finally, the tenacious adherence to the diagnosis of paralytic poliomyelitis may also be due to an intense desire to hold fast to a heroic image of FDR. In that respect, Roosevelt overcame his permanent paralysis to rise to the Presidency of the USA. His impact on world history through his leadership during the Great Depression and the Second World War and the creation of the organization⁵³ that spearheaded the development of poliovirus vaccines that stemmed the epidemics of poliomyelitis⁵⁴ remains admirable regardless of the cause of his illness.

These possible explanations are admittedly subjective. By contrast, the clinical findings, a consideration of the pathogenesis of the illness and the accompanying statistical analyses objectively indicate that GBS was the most probable cause of FDR’s neurological disease.

Footnotes

Acknowledgements

We thank Julie M Trumble, Anne B Howard and Alexander Bienkowski from the Moody Memorial Medical Library from the University of Texas Medical Branch for finding key references concerning the CSF abnormalities in paralytic poliomyelitis and Guillain–Barré syndrome. We also thank Jessica B Murphy from the Boston Medical Library in the Francis A Countway Library of Medicine for the copy of the telegram that Dr Robert W Lovett sent to Dr William W Keen.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author biographies

Armond S Goldman, MD, is Emeritus Professor of Pediatrics at the University of Texas Medical Branch in Galveston, Texas. His research concerns the immune system in human milk, ontogeny of immunity, the evolution of the mammary gland, and genetic immunodeficiencies.

Elisabeth J Schmalstieg, MD, is a neurological Consultant in Texas City, Texas and a Clinical Assistant Professor of Neurology at the University of Texas Medical Branch. She has a great deal of experience in the diagnosis and management of many types of neurological diseases, including the inflammatory and metabolic neuropathies.

Charles F Dreyer, MD, is Associate Professor of Neurology and Pediatrics at the University of Texas Medical Branch. He is a paediatric neurologist with a wide experience in neurological disorders.

Frank C Schmalstieg, MD, PhD, is Professor of Pediatrics at the University of Texas Medical Branch in Galveston, Texas. His research concerns mechanisms of congenital and genetic immunodeficiencies, including neutrophil defects and combined immunodeficiencies. He has conducted in-depth studies concerning the molecular mechanisms of neutrophil function and of inflammatory lung disease.

Daniel A Goldman, MD MPH, is an epidemiologist and software developer in Seattle, Washington. He has developed a large-scale software framework (Vitalnet) that lets users analyse health-related data sets such as birth, hospital and mortality data. He has also carried out research on the appropriate use of age-adjusted rates and standardized mortality rates in public health practice.

References

Goldman

Schmalstieg

Freeman

Jr . What was the cause of Franklin Delano Roosevelt’s paralytic illness? Journal of Medical Biography 2003; 11: 232–240.

Dietz

Lezana

Garcia Sancho

. Predictors of poliomyelitis case confirmation at initial clinical evaluation: implications for poliomyelitis eradication in the Americas. International Journal of Epidemiology 1992; 21: 800–806.

Alcala

. The differential diagnosis of poliomyelitis and other acute flaccid paralyses. Boletin Medico del Hospital Infantil de Mexico 1993; 50: 136–144.

Mulder

. Clinical observations on acute poliomyelitis. Annals of the New York Academy of Sciences 1995; 753: 1–10.

Jubelt

Miller

Viral infections. In: Rowland

(ed). Merritt’s Textbook of Neurology, (9th ed.). Baltimore: Williams & Wilkins, 1995, pp. 145–147.

Marx

Glass

Sutter

. Differential diagnosis of acute flaccid paralysis and its role in poliomyelitis surveillance. Epidemiologic Reviews 2000; 22: 298–316.

Kidd

Manji

Brown

. Acute paralytic illness. Postgraduate Medical Journal 1996; 72: 699–701.

Löffel

Rossi

Mumenthaler

. The Landry–Guillain–Barré syndrome. Complications, prognosis and natural history in 123 cases. Journal of the Neurological Sciences 1977; 33: 71–79.

Samantray

Johnson

Mathai

. Landry–Guillain–Barré–Strohl syndrome: a study of 302 cases. Medical Journal of Australia 1977; 2: 84–91.

10.

Asbury

Cornblath

. Assessment of current diagnostic criteria for Guillain–Barré syndrome. Annals of Neurology 1990; 27: S21–S24.

11.

Ropper

. The Guillain–Barré syndrome. New England Journal of Medicine 1992; 326: 1130–1136.

12.

Hughes

Rees

. Clinical and epidemiologic features of Guillain–Barré syndrome. Journal of Infectious Diseases 1997; 176: S92–S98.

13.

Black

. Franklin Delano Roosevelt: Champion of Freedom, London: George Weidenfeld & Nicolson, 2003.

14.

Coker

. Franklin D. Roosevelt: A biography, Westport, Connecticut: Greenwood Press, 2005.

15.

Winkler

. Franklin D. Roosevelt and the making of modern America, New York: Pearson Educational, Inc, 2006.

16.

Smith

. FDR, New York: Random House, 2007.

17.

Brands

. Traitor to His Class: The privileged life and radical presidency of Franklin Delano Roosevelt, New York: Doubleday, 2008.

18.

Shell

. Polio and its aftermath: The paralysis of culture, Cambridge, MA: Harvard University Press, 2005.

19.

Oshinsky

. Polio. An American story, New York: Oxford University Press, Inc., 2005.

20.

Wilson

. Living with polio: The epidemic and its survivors, Chicago, IL: University of Chicago Press, 2005.

21.

Peters

. Epidemic: The battle against poliomyelitis, New York: Benchmarks Books, 2005.

22.

Brenner

Castellani

Gabaree

. Immune changes in humans during cold exposure: effects of prior heating and exercise. Journal of Applied Physiology 1999; 87: 699–710.

23.

Parry

Steinberg

. Guillain-Barré syndrome: From diagnosis to recovery, New York: Demos Medical Publishing, 2007.

24.

Friedenberg

. Franklin D. Roosevelt: His poliomyelitis and orthopaedics. Journal of Bone and Joint Surgery (American) 2009; 91: 1806–1813.

25.

Lomazow S. My-paper-on-fdr-in-journal-of-medical.html, Available from http://fdrsdeadlysecret.blogspot.com/ (2010, accessed 22 August 2010).

26.

Case of Franklin Delano Roosevelt, Telegram from Lovett RW to Keen WW. 24 August 1921 (BMS c39). , Boston, MA: Boston Medical Library in the Francis A. Countway Library of Medicine.

27.

Case of Franklin Delano Roosevelt. Letter from Lovett RW to Draper G, 12 September 1921. (BMS c39). Boston, MA: Boston Medical Library in the Francis Countway Library of Medicine.

28.

Gallagher

. FDR’s splendid deception, New York: Dodd, Mead, 1985.

29.

Roosevelt E. Letter to James Roosevelt, 14 August 1921. Hyde Park, New York: Franklin D Roosevelt Library and Museum.

30.

Witte

Henderson

. The cerebrospinal fluid in type 3 poliomyelitis. American Journal of Epidemiology 1966; 83: 189–195.

31.

Link

Wahren

Norrby

. Pleocytosis and immunoglobulin changes in cerebrospinal fluid and herpesvirus serology in patients with Guillain-Barré syndrome. Journal of Clinical Microbiology 1979; 9: 305–316.

32.

Landry

. Note sur la paralysie ascendante aigue. Gazette Hebdomadaire de Médecine et de Chirurgie 1859; 6: 472–474, 486-488.

33.

Lovett

. The treatment of infantile paralysis, Philadelphia, PA: Blakiston’s Son & Co, 1916.

34.

Guillain

Barré

Strohl

. Sur un syndrome de radiculo-nevrite hyperalbuminose du liquide cephalorachidien sans reaction cellulaire. Bulletins et Memoires de la Societe Medicale des Hopitaux de Paris 1916; 40: 1462–1470.

35.

Leyden E

von

. Über poliomyelitis und Neuritis. Zeitschrift für klinische Medizin 1879/80; 472: 486–493.

36.

Racaniello

. One hundred years of poliovirus pathogenesis. Virology 2006; 344: 9–16.

37.

Peterson

. Clinical aspects of Campylobacter jejuni infections in adults. Western Journal of Medicine 1994; 161: 148–152.

38.

Islam

Jacobs

van Belkum

. Axonal variant of Guillain-Barré syndrome associated with Campylobacter infection in Bangladesh. Neurology 2010; 74: 581–587.

39.

Ruts

Drenthen

Jongen

. Pain in Guillain–Barré syndrome: a long-term follow-up study. Neurology 2010; 75: 1439–1447.

40.

Melillo

Sethi

Mohsenin

. Guillain–Barré syndrome: rehabilitation outcome and recent developments. Yale Journal of Biology and Medicine 1998; 71: 383–389.

41.

Ropper

. Severe acute Guillain–Barré syndrome. Neurology 1986; 36: 429–432.

42.

Sakakibara

Uchiyama

Kuwabara

. Prevalence and mechanism of bladder dysfunction in Guillain–Barré Syndrome. Neurourology and Urodynamics 2009; 28: 432–437.

43.

Kuwabara

Yuki

Koga

. IgG anti-GM1 antibody is associated with reversible conduction failure and axonal degeneration in Guillain–Barré syndrome. Annals of Neurology 1998; 44: 202–208.

44.

Kaida

Kusunoki

. Antibodies to gangliosides and ganglioside complexes in Guillain-Barré syndrome and Fisher syndrome: Mini-review. Journal of Neuroimmunology 2010; 223: 5–12.

45.

Lopez

Zhang

. Passive transfer of IgG anti-GM1 antibodies impairs peripheral nerve repair. Journal of Neuroscience 2010; 30: 9533–9541.

46.

Bayes

. An essay towards solving a problem in the doctrine of chances. Philosophical Transactions of the Royal Society of London 1763; 53: 370–418.

47.

Stigler

. The history of statistics: The measurement of uncertainty before 1900, Cambridge, MA: Belknap Press of Harvard University Press, 1986.

48.

Research criteria for diagnosis of chronic inflammatory demyelinating polyneuropathy (CIDP): report from an Ad Hoc Subcommittee of the American Academy of Neurology AIDS Task Force. Neurology 1991; 41: 617-618.

49.

Haq

Fries

Pendlebury

. Chronic inflammatory demyelinating polyradiculoneuropathy. A study of proposed electrodiagnostic and histologic criteria. Archives of Neurology 2000; 57: 1745–1750.

50.

Darley

Gross

A hypothesis-confirming bias in labelling effects. In: Stangor

(ed). Stereotypes and prejudice: Essential readings, East Sussex, UK: Psychology Press, 2000.

51.

Bacon

Novum organum. Original work published in 1620. In: Burtt

(ed). The English Philosophers from Bacon to Mill, New York: Random House, 1939, pp. 24–123.

52.