The Etiology of Rheumatoid Arthritis

by William McKendree Jefferies MD

Author of: THE SAFE USES OF CORTISOL. (1996).
Originally published as SAFE USES OF CORTISONE. (1981).

Medical Hypotheses
Volume 51, Issue 2. Pages 111-114.
(August 1998)

Received: April 7, 1997
Accepted: May 13,1997

– – Abstract – –

The etiology of rheumatoid arthritis has been elusive, but it finally seems to
be explained by a combination of three factors:

(i) a relatively mild deficiency of cortisol, the normal adrenocortical hormone
that is essential for normal immunity but which has achieved a bad reputation
because of the use of excessive dosages of it or its stronger derivatives,

(ii) a deficiency of dehydroepiandrosterone (DHEA), the chief androgen produced
by the human adrenal cortex but which has been little studied, and

(iii) infection by organisms such as mycoplasma, which have a relatively low
virulence, are difficult to culture in the laboratory, and cause inflammation
and destruction of tissue in periarticular and articular areas of
immunocompromised hosts.

The mild adrenocortical deficiency apparently is sufficient to impair immunity,
especially after stress, and permit these organisms to cause inflammatory

Further studies are necessary to determine optimum therapy, but it will probably
include safe physiologic dosages of cortisol and DHEA plus antibiotic treatment
of infection by mycoplasma or other causative organisms.

– – –

After Hench and his associates reported the dramatic beneficial effects of
adrenocorticotropic hormone (ACTH) and of cortisone in patients with rheumatoid
arthritis in 1949 (1), the possibility that rheumatoid arthritis might result
from a deficient production of cortisone was considered, but tests of
adrenocortical and pituitary function at that time were limited.

In 1951 (2) Konrad Dobriner, a pioneer steroid chemist, reported that studies of
24 h urine collections of five patients with rheumatoid arthritis (3 men and 2
women) revealed that they excreted steroid hormone metabolites but at a
consistently low normal level and he concluded that ‘there is presumed to be a
decreased hormone formation by the adrenal glands in rheumatoid arthritis’.

Nine years later, Kelley and Ely (3), in a review of the production and
metabolism of adrenocorticosteroids in connective tissue disease, summarized by

‘It appears that neither adrenocortical function nor steroid metabolism is
completely normal in patients with connective tissue disease, but the
abnormalities that exist may be quite subtle and difficult to demonstrate

As an endocrinologist I do not have an opportunity to study and treat many
patients with rheumatoid arthritis, but in 1967 (4) I described two teenage
girls with juvenile rheumatoid arthritis, one of whom at age 13 shortly after
the menarche developed migratory pains in the hips, elbows, metatarsal and
temporo-mandibular joints. After she had been unable to attend school for
several weeks, studies in Western Reserve University Hospital resulted in a
diagnosis of probable rheumatoid arthritis.

Fractionation of urinary steroids showed they were within low normal range
except for very low excretion of dehydroepiandrosterone (DHEA).

Serum 17-hydroxy-corticosteroid levels (an indirect measurement of circulating
levels of cortisone and hydrocortisone) were also in low normal range, but her
symptoms cleared within 2 weeks after she began taking 5 mg cortisone acetate
four times daily, before meals and at bedtime, a dosage that was only one half
of a normal replacement dosage for an adrenalectomized patient, or one half of
the amount produced by a person with normal adrenals under unstressed

She remained free of symptoms as long as she continued this dosage, but symptoms
returned at varying intervals of 2 weeks to 4 months on four occasions during
subsequent years when her dosage was decreased to 2.5 mg four times daily,
clearing when the dosage was returned to 5 mg four times daily.

Six months after cortisone therapy had been started, a small non-toxic
enlargement of her thyroid gland was noted, and the addition of triiodothyronine
[thyroid T3], 25-mcg by mouth daily, resulted in a return of her thyroid gland
to normal size. She was married at age 19 and had normal pregnancies at age 20
and 22, cortisone and triiodothyronine therapy being continued at the same
dosages through both pregnancies.

At age 22 her latex fixation test was positive, but erythrocyte sedimentation
rate was normal (8 mm per hour, corrected). After it was reported that cortisone
must be converted to cortisol before having its physiologic effects, at age 24
her steroid therapy was changed to cortisol, 2.5 mg four times daily.

She was asymptomatic on this dosage for several months, but then arthritis
symptoms returned, so the dosage was increased to 5 mg four times daily and
symptoms again cleared.

At age 27 sedimentation rate and SMAC were normal, and an ACTH test revealed a
baseline plasma cortisol at 9 a.m. of 28 ~tg/dl, rising to 35 ~tg/dl after
Cortrosyn, consistent with low adrenal reserve.

It was therefore evident that 5 mg of cortisone or cortisol four times daily,
which was just half of a replacement dosage, was sufficient to maintain an
apparently complete remission in her arthritis, whereas one half of that dosage
was not.

It was also evident that this dosage did not interfere with her having two
normal pregnancies.

The other teenager was first seen at age 14 when she developed both rheumatoid
arthritis and hyper-thyroidism with a diffuse goiter. Her arthritis, which was
most severe in the knees, wrists and elbows, with pain, swelling and limitation
of motion, had been treated with aspirin with slight improvement, but it
subsided completely when she was given cortisone acetate, 5 mg every 8 hours,
and her hyperthyroidism cleared with propylthiouracil treatment.

After dis-continuance of the propylthiouracil but while she continued taking
cortisone, 5 mg every 8 hours, both auto-immune disorders remained in remission
for the 7 years that I was able to follow her.

During this time she had a normal menarche [first menstruation], grew from 67
1⁄4 inches to 69 3⁄4 inches (168 to 174 cm) in height, graduated
from high school and college, and began teaching in an elementary school without
any relapse in her arthritis or hyperthyroidism, so there was no indication that
this therapy interfered with her normal growth or development.

Fractionation of her urinary steroids also revealed a relatively low excretion
of DHEA.

In that report (4) I also presented evidence that five women with rheumatoid
arthritis had diminished excretion of dehydroepiandrosterone (DHEA), a normal
adrenal androgen, and stated that ‘these observations suggest that an
abnormality in steroid metabolism may exist in rheumatoid arthritis’.

This abnormality consisted of a mild deficiency of production of cortisol with
an accompanying deficiency of production of DHEA.

In 1984 Masi and his associates (5) reported that DHEA excretion of eight women
with rheumatoid arthritis was significantly lower after metyrapone
administration than that of normal women, supporting the impression that women
with rheumatoid arthritis have diminished production of DHEA.

More recent reports have confirmed that a deficient production of cortisol,
especially at times of greatly increased stress, occurs in patients with
rheumatoid arthritis.

Neeck et al in 1990 (6) reported that patients with rheumatoid arthritis had
abnormal diurnal patterns of blood levels of cortisol and that the degree of
abnormality varied with the degree of severity of their arthritis as manifested
by their sedimentation rates as well as by their arthritic symptoms.

Chikanza and his associates in 1992 (7) reported evidence that patients with
rheumatoid arthritis had a defective hypothalamus-pituitary-adrenal (HPA)
response as evidenced by a diurnal pattern of cortisol secretion at the lower
limit of normal in contrast to patients with osteomyelitis, and a failure to
increase cortisol secretion following surgery despite high levels of
interleukin-1 beta and interleukin-6, two cytokines that normally appear to
regulate stimulation of the HPA axis.

These reports suggested that a defective stimulation of the adrenals through the
HPA axis might result in relative cortisol deficiency despite potentially normal
pituitary and adrenocortical function.

In 1994 Hall et al (8) reported that patients with untreated rheumatoid
arthritis had significantly increased levels of ACTH, without accompanying
elevation of cortisol, consistent with adrenal hypo- responsiveness, and in 1996
Gudbj~Srnsson et al (9) reported an impaired cortisol response with normal ACTH
secretion in patients with active rheumatoid arthritis, also consistent with
relative adrenocortical deficiency.

Although these reports varied in their approaches, they all indicated that a
deficiency of cortisol production, especially in response to stress, a
deficiency that might be relative or absolute, and at least sometimes associated
with a deficient production of DHEA, occurred in patients with rheumatoid
arthritis, but they did not explain why everyone with adreno-cortical deficiency
did not develop rheumatoid arthritis, nor did they explain why the disease was
manifested by inflammation and destruction of joints.

Hence an additional factor or factors must be involved.

An explanation for these aspects of the disease was provided by T. McPherson
Brown, a rheumatologist who died in 1989 after spending many years accumulating evidence that rheumatoid arthritis is associated with an inflammatory infection in and around the joints produced by a bacteria-like organism termed mycoplasma (10).

This organism has a propensity to grow and cause inflammation in periarticular
areas under circumstances in which the host’s immune system is impaired, and it
can be controlled by pro-longed administration of a suitable antibiotic.

Brown stated that ‘The length of time patients require in treatment can vary
widely, depending primarily on how long they have had the disease. In most
entrenched and recalcitrant cases, it can take up to thirty months from the
beginning of therapy until the patient clearly turns the corner toward
improvement, and the achievement of a lasting remission can take several years’

He also stated that prednisone in suitable small dosages (1 to 10 mg daily) was
helpful in some cases, but he did not mention studies of adrenocortical
function, and he did not explain why patients with rheumatoid arthritis were
more susceptible to the development of infections with mycoplasma.

This susceptibility could be explained by the evidence reported by numerous
investigators over the past forty years that physiologic amounts of cortisol are
necessary for the development and maintenance of normal immunity (12), so a
deficiency of cortisol could impair immunity and probably provide fertile ground
for infection with mycoplasma.

Hence at least three factors seem to contribute to the development of rheumatoid

An inadequate adrenocortical production of cortisol, either relative or
absolute, is probably a necessary factor in all cases since all seem to be
associated with mild adrenocortical deficiency, either primary in the adrenals
or secondary to deficient production of ACTH (factor #1).

The mild deficiency of cortisol might not only lower resistance to stress but
also depress immunity and lower resistance to a micro-organism such as
mycoplasma that may cause peri-articular inflammation and joint destruction in
immunocompromised hosts (factor #2).

Thirdly, the decreased production of the adrenocortical androgen, DHEA, might
explain the predilection of rheumatoid arthritis for women as well as contribute
to the impairment of immunity and to the impairment of healing of inflamed and
damaged tissues (factor #3).

Any or all of these factors might be inherited, explaining the familial tendency
for development of the disease.

It would therefore be advisable for patients with rheumatoid arthritis to have
baseline studies of adrenocortical and pituitary function (plasma cortisol,
ACTH, and DHEA sulfate, and a Cortrosyn stimulation test), as well as tests to
determine the nature and activity of the disease (erythrocyte sedimentation
rate, complete blood count with differential, SMA, ANA, rheumatoid factor, and
any other studies that might be indicated in the specific patient), and then be
treated with safe, physiologic dosages of cortisol (4,13) for the rest of their
lives, and with dosages of a suitable antibiotic for a sufficient period to
clear infection with mycoplasma or any other organism that might be causing
inflammation and damage to peri- articular tissues.

In evaluating tests of adrenocortical function, it should be remembered that
normal ranges for these tests were determined before mild adreno-cortical
deficiency was recognized and hence might be too broad.

If plasma DHEA-sulfate or urinary DHEA excretion is low, supplementation with
DHEA by mouth in a dosage sufficient to bring plasma levels to normal should
also be helpful.

A starting dose of 5 mg DHEA four times daily, before meals and at bedtime,
seems to be satisfactory. If arthritic symptoms are severe, a larger dosage of
cortisol might produce improvement and relief more quickly, but once relief
occurs, the dosage should be tapered to a physiologic maintenance level while an
antibiotic is administered to control the infection caused by the mycoplasma.

It might be questioned whether any of the more widely used stronger derivatives
of cortisol or cortisone, such as prednisone, triamcinolone, etc., might be
administered instead of cortisol, but nature usually has a reason for her choice
of normal hormones, so it seems wise to follow her example as closely as
possible, on a schedule that mimics her production as closely as is feasible,
especially for long-term use.

Experience with the two girls with juvenile rheumatoid arthritis suggests that
in its early stages rheumatoid arthritis might be controlled by administration
of physiologic dosages of cortisol without additional DHEA or an antibiotic, so
that possibility should be investigated.

It is possible that the combination of decreased cortisol and decreased DHEA is
necessary to lower immunity sufficiently to cause rheumatoid arthritis and that
giving adequate doses of cortisol or DHEA or both may prevent it, but the two
girls’ responses to treatment suggest that adequate cortisol will prevent
rheumatoid arthritis even though DHEA remains low.

Nevertheless, it seems advisable to correct any abnormality as much as possible
in order to obtain optimum health, so correction of any abnormality found should
be the safest course.

The occurrence of autoimmune thyroid disorders in these two patients also is
consistent with evidence that other autoimmune disorders may be associated with
mild adrenocortical deficiency and hence might also benefit from physiologic
dosages of cortisol (14,15), so studies of pituitary and adrenocortical function
in patients with other autoimmune disorders should be helpful.

Because susceptibility to these disorders also may be inherited, family members
might benefit from studies of their HPA axis and treatment of any adrenocortical
deficiency before it contributes to the development of one of these disorders.

Treatment with safe, physiologic dosages of cortisol and DHEA should probably be
continued throughout patients’ lives, as is advisable for other patients with
mild adrenocortical deficiency, with temporary increases in dosages during times
of increased stress (13).

Because the harmful side effects that have given treatment with large,
pharmacologic dosages of glucocorticoids such a bad reputation do not occur with
physiologic dosages of the normal hormone, the effectiveness and safety of
treatment with physiologic dosages of cortisol having been documented by over
two thousand patient-years of treatment (12), this therapeutic approach will, it
is hoped, not be impeded further by the failure of pharmaceutical companies to
promote or even mention new uses of safe dosages of cortisol.

This peculiar behavior apparently results from a government regulation that
states that when a new use is found for a medication whose patent has expired it
must receive therapeutic trials to prove its safety for this new use before the
new use can be advertised.

Such studies are obviously important, but they are also expensive, and companies
are naturally hesitant to undertake such studies without the protection provided
by a patent.

Other hormones have apparently never had this problem, but they have not had a
bad reputation like cortisol nor the large, relatively untapped potential for
its use.

Hence, one hopes, these regulations will be reconsidered and revised to permit
advertisement of this normal hormone in physiologic dosages for patients who are
deficient in it.

– – Acknowledgment – –

The author is indebted to Mrs Patricia Shannon for help in obtaining references
for this manuscript.

– – References – –

1.) Hench PS, Kendall EC, Slocumb CH, Polley HE.The effect of a hormone of the
adrenal cortex (17-hydroxy-11-dehydro-corticosterone: Compound E) and of
pituitary adrenocortico-tropic hormone on rheumatoid arthritis: preliminary
report. Proc Staff Meet Mayo Clin 1949; 24: 181-197.

2.) Dobriner K. Adrenal function in rheumatoid arthritis. Bull Rheumatic
Diseases II 1951: 23-24.

3.) Kelley V C, Ely R S. Production and metabolism of adreno-corticosteroids in
connective tissue disease. Ann NY Acad Sci 1960; 86: 1115-1128.

4.) Jefferies W McK. Low dosage glucocorticoid therapy. Arch Intern Med 1967;
119: 265-278.

5.) Masi A T, Josipovic D B, Jefferson W B. Low adrenal andro-genic-anabolic
steroids in women with rheumatoid arthritis (RA): gas-liquid chromatographic
studies of RA patients and matched control women indicating decreased l
l-deoxy-17- ketosteroid excretions. Semin Arthritis Rheum 1984; 14: 1-23.

6.) Neeck G, Federlin K, Graef V, Rusch D, Schmidt KL. Adrenal secretion of
cortisol in patients with rheumatoid arthritis. J Rheumatol 1990; 17: 24-29.

7.) Chikanza I C, Petros P, Kingsley G E, Chrousos G, Panayi G S. Defective
hypothalamic response to immune and inflammatory stimuli in patients with
rheumatoid arthritis. Arthritis Rheum 1992; 35: 1281-1288.

8.) Hall J, Morand EF, Medback S et al. Abnormal hypo-
thalamic-pituitary-adrenal axis function in rheumatoid arthritis. Arthritis
Rheum 1994; 37:1132-I 137.

9.) Gudbj Srnsson B, Skogseid B, Oberg K, Weid L, H~illgren R. Intact
adrenocortico-tropic hormone secretion but impaired cortisol response in
patients with active rheumatoid arthritis: effect of glucocorticoids. J
Rheumatol 1996; 23: 596-Rheum 1954; 14: 1-23.

10.) Clark H W, Coker-Vann M R, Bailey J S, Brown T McP. Detection of
mycoplasmal antigens in immune complexes from rheumatoid arthritis synovial
fluids. Ann Allergy 1988; 60: 394-398.

11.) Brown T McP, Scammell H. The road back: rheumatoid arthritis: its cause and
its treatment. In: ScammeU H. The Arthritis Breakthrough. New York: Evans, 1993:

12.) Jefferies W McK. Cortisol and immunity. Med Hypotheses 1991; 34: 198-208.

13.) Jefferies W McK. Safe Uses of Cortisol, 2nd edn. Springfield, IL: Thomas,
1996: Chapter 4, p. 35-66; Chapter 6, p. 91-104.

14.) Jefferies W McK. Mild adrenocortical deficiency, chronic allergies,
autoimmune disorders and the chronic fatigue syndrome: a continuation of the
cortisone story. Med Hypotheses 1994; 42: 183-189.

15.) Jefferies W McK. Safe Uses of Cortisol, 2nd edition. Springfield, IL:
Thomas, 1996: Chapter 8, p. 114-127.