Anxiety or Caffeinism: A Diagnostic Dilemma

BY JOHN F. GREDEN, MI).

inquiries about drugs that are less frequently used are

The author reports that high intake ofcaffeine (‘ ‘caffein-
characteristic.
ism ‘) can produce
‘ symptoms that are indistinguishable
This paper reviews the pharmacological and medical
f rom those ofanxiety neurosis, such as nervowness, irri-
literature on the effects of caffeinism and presents three
tability, tremulousness, occasional muscle twitchings, in-
clinical case reports to emphasize that occasional cogni-
somnia, sensory dist urbances tachypnea.
, palpitations,
zance must be given to the use of coffee, tea, and other
flushing, arrhythmias, diuresis, and gastrointestinal dis-
sources of caffeine in research, clinical evaluation, and
turbances. The caffeine withdrawal syndrome and the
routine treatment of anxious patients.
headache associated with it may also mimic anxiety.
Patients with caffeinism will generally be identified only
by routine inquiry into their caffeine intake. The psychia- PHARMACOLOGY
trist should especially suspect caffeinism in patients who
do not respond to psychopharmacological agents or who
Caffeine, theophylline, and theobromine pharmacolog-
have psychophysiological complaints and rec urrent head-
ically impact most physiological systems (6, 7). With re-
aches, chronic coffee-drinkingpatients on inpatient psy-
gard to the central nervous system (CNS), caffeine is a
chiatric services, and “hyperkinetic” children. Three case
powerful stimulant. Indeed, the popularity of caffeinated
reports illustrate the syndrome.
beverages is undoubtedly due to such stimulation. It is
claimed that coffee and tea promote rapid, clear think-
ing, improved intellectual effort, enhanced mental
IN THEIR LIST of diagnostic criteria for adult psychiatric acuity, and decreased drowsiness, fatigue, and reaction
illness, Feighner and associates ( I ) stipulated that before time-all valued commodities in an achievement-
the diagnosis of anxiety neurosis can be assigned, the ab- oriented society. Caffeine affects all parts of the cortex,
sence of any medical illness that could account for symp- accounting for most of these stimulating actions. Respi-
toms of anxiety must first be determined. Unfortunately, ratory, vasomotor, and vagal centers of the medulla
anxiety looks and feels the same, regardless of etiology. are also affected by caffeine, as are all parts of the spinal
The task of diagnosing it accurately can be a difficult cord at high doses (6-8).
one (2-4). Caffeine’s action on the cardiovascular system can be
Relevant to this endeavor is the overlooked fact that detected in systemic, coronary, and cerebral circulation
high doses of caffeine-or “caffeinism”-can produce patterns. The coronary arteries and the pulmonary and
pharmacological actions that cause symptoms essentially general systemic vessels become dilated following caf-
indistinguishable from those of anxiety neurosis (5). The feine ingestion. Heart rate and force of contraction also
consumption of coffee and tea is ubiquitous in the United increase. Many of these peripheral effects cannot be ob-
States. At least one billion kilograms of coffee are con- served directly, however, because they tend to be masked
sumed annually (6, p. 354), yet a review of the pertinent by central stimulating effects upon the medulla. Thus, at
literature failed to reveal any significant mention of caf- moderate doses an individual may notice no change
feinism. Furthermore, a random review of 100 recent out- in heart rate, no tachycardia, or even a slight bradycar-
patient psychiatric records at my medical center failed to dia (6, 7).
reveal a single listing of coffee- or tea-drinking patterns, Within the gastrointestinal system, caffeine signifi-
despite the fact that 42 of these records referred to anx- cantly affects gastric secretion, resulting in prolonged
iety symptoms. Commonly employed mental status augmentation of both volume and acidity. The renal sys-
tem is also affected, producing the commonly experi-
forms-both free-form and operationally defined-make
no provisions for noting caffeine consumption, although enced “coffee diuresis.” Finally, caffeine relaxes smooth
muscles (especially of the bronchi), strengthens the con-
traction of skeletal muscles, and increases basal meta-
Read at the 127th annual meeting of the American Psychiatric Associa- bolic rate an average of 10 percent among regular coffee
tion, Detroit, Mich., May 6-10, 1974. drinkers (6, 7).
At the time this paper
was written, Dr. Greden was Director of Psychi- Dosage is a relevant consideration in the study of caf-
atric Research, Walter
Reed Army Medical Center, Washington, D.C.
He is currently Assistant Professor of Psychiatry, University of Michi- feine pharmacology. Fifty to 200 mg. of caffeine is typi-
gan Medical Center, Ann Arbor, Mich. 48104. cally required to produce expected pharmacological ac-

Am J Psychiatry /31.10. October 1974 1089

ANXIETY OR CAFFEINISM

TABLE 1 especially among people not accustomed to caffeine at

Some Common Sources of Caffeine (6. 7. 9. 10) bedtime. This latter finding may correlate with the debat-
able development of tolerance among heavy coffee drink-
Approximate Amounts of ers (13). Caffeine also seems to counteract the sedative/
Source Caffeine Per Unit hypnotic effect of medications. When caffeine was taken
together with pentobarbital at bedtime in one study, the
Beverages effect was approximately the same as that of a pla-
Brewed coffee 100-150 mg. per cup
Instant coffee 86-99 mg. per cup
cebo (14).
Tea 60-75mg.percup In addition to toxic manifestations of caffeine stimu-
Decaffeinated coffee 2-4 mg. per cup lation, similar clinical complaints can result from a char-
Cola drinks 40-60 mg. per glass acterislic caffeine withdrawal syndrome. !n a survey
Prescription medications
among housewives by Goldstein and Kaizer (15), heavy
APCs (aspirin, phenacetin. caffeine) 32 mg. per tablet
Cafergot 100 mg. per tablet coffee drinkers described a typical set of dysphoric symp-
Darvon compound 32 mg. per tablet toms ifthey omitted their morning coffee. Symptoms in-
Fiorinal 40 mg. per tablet cluded irritability, inability to work effectively, nervous-
Migral 50 mg. per tablet ness, lethargy, and restlessness. Of special interest,
Over-the-counter analgesics
headache was also a frequent complaint among those
Anacin, aspirin compound, Bromo
Seltzer 32 mg. per tablet withdrawing from caffeine. There is little doubt that caf-
Cope. Easy-Mens, Empirin compound, feine withdrawal headache is a real phenomenon since it
Midol 32 mg. per tablet has been produced experimentally. In 1943 Dreisbach
Vanquish 32 mg. per tablet
and Pfeiffer (16) described this caffeine withdrawal head-
Excedrin 60 mg. per tablet
ache as being “remarkably constant, beginning with leth-
Pre-Mens 66 mg. per tablet
Many over-the-counter cold preparations 30 mg. per tablet argy on the morning ofthe first day followed by a feeling
Many over-the-counter stimulants 100 mg. per tablet of cerebral fullness about noon and an actual headache
beginning in the afternoon.” Although initially localized,
headaches usually became generalized and throbbing and
were sometimes associated with nausea, rhinorrhea, de-
tions. Drill’s Pharmacology (7) refers to doses of over pression, drowsiness, yawning, and a disinclination to
work. Not surprisingly, caffeine withdrawal headaches
250 mg. as being “large.” Of clinical significance, this
dosage is frequently exceeded in everyday life. As in- respond to caffeine but tend to recur the next
dicated in table 1, caffeine is widely available in a multi- day (16, 17).
plicity of beverages and medications. When cumulatively Toxic manifestations are not restricted soley to the
considered, an individual can ingest doses much higher CNS, of course. Varying with the dosage, the degree of
individual tolerance, and the existing autonomic condi-
than 250 mg.-often without being aware of it. For ex-
ample, three cups of coffee, two over-the-counter head- tions within the cardiovascular system. caffeine can
ache tablets, and one cola drink consumed in one morn- cause palpitations, extrasystoles, tachycardia, arrhyth-
mias, flushing, and, in extremely large doses, even
ing approximate 500 mg. of caffeine intake. Among
heavy coffee or tea drinkers, dosages frequently exceed marked hypotension and circulatory failure (18, 19).
this by gross amounts. Within the gastrointestinal system, caffeine has the po-
tential to produce nausea, vomiting, diarrhea, epigastric
pain, and occasionally, peptic ulcer and hematemesis.
These actions may result from pharmacological en-
ANXIETY MANIFESTATIONS OF CAFFEINISM hancement of gastric secretions or simply from the effect
of irritating oils included in the beverage (7). Miscel-
Symptoms of caffeinism are best understood as dose- laneous toxic manifestations of caffeine include reported
related extensions ofcaffeine’s expected pharmacological cases ofdehydration, fever, and edema (5, 20).
actions. Listing these symptoms emphasizes that the syn- Obviously, if a combination of these symptoms caused
drome ofcaffeinism reads like a classic description of an by caffeinism was reported by a patient, the constellation
anxiety attack. For example, within the CNS, toxic could easily cause clinical confusion. Three case reports
symptoms referable to stimulation include “nervous- illustrate this point.
ness,” irritability, agitation, headache, tachypnea, tremu-
lousness, reflex hyperexcitability, and occasional muscle
twitchings (6, 11). (The mass media have aptly described
CASE REPORTS
this constellation as “coffee nerves” and advocate
“morning rationing” as an appropriate treatment for af-
Case I. A 27-year-old nurse who was married to an Army
flicted subjects.) Toxic sensory disturbances may include physician applied for evaluation at a military outpatient medi-
hyperesthesia, ringing in the ears, and visual flashes of cal clinic because of lightheadedness, tremulousness, breath-
light (6, 11). Insomnia is also a frequent consequence of lessness, headache, and “irregular heartbeat” occurring sporad-
the CNS stimulation secondary to caffeinism (12). Both ically approximately two or three times a day. The symptoms
delayed sleep onset and frequent breaks in sleep occur, had developed gradually over a three-week period. She denied

1090 Am J Psychiatry 131:10. October /974

 JOHN F. GREDEN

precipitating stresses. When the evaluating physician corn- after his initial visit he reported distinct improvement of his
mented on her apparent anxiety, she admitted being apprehen- long-standing tremulousness, loose stools, and insomnia. His
sive but correlated it with the presence of palpitations, chest job apprehension continued unabated, but he cynically noted
discomfort, and irregular heartbeat. Physical examination, a that he “was still working for the same SOB.” To reinforce a
multitude of laboratory studies (sodium, potassium, bicarbo- cause-and-effect relationship, he too was “challenged” with
nate, chloride, calcium, phosphorous, uric acid, alkaline large doses ofcaffeine for several days in succession. He experi-
phosphatase. lactate dehydrogenase blood urea nitrogen, T-3 enced a prompt recurrence of symptoms until the pattern of
and T-4 thyroid, and glucose tolerance tests) were all within consuming large doses of caffeine again ceased. Scores on the
normal limits. An electrocardiogram (EKG) revealed charac- Hamilton Anxiety Scale obtained three months later were sig-
teristic premature ventricular contractions (PVCs). nificantly lower.
During her final session with the evaluating internist, she was
given quinidine sulfate and instructed on its use should her Case 3. A 34-year-old Army personnel sergeant was referred
PVCs fail to disappear. She was also referred to the psychiatric for psychiatric evaluation and treatment of severe, recurrent
outpatient clinic with an intriguing diagnosis: “anxiety reaction tension headaches. Three thorough medical evaluations within
(probably secondary to the fear that her husband would be two years, including a complete hypertensive workup, had been
transferred to Viet Nam).” However, the patient refused to ac- negative. The referring physician noted that the patient was
cept this determination despite continuation of her symptoms. “driving them [the staff] crazy with frequent clinic visits.” The
She diligently searched for a dietary cause (“Perhaps it is hy- doctor was also overtly concerned about the patient’s excessive
poglycemia”). After approximately 10 days, she convincingly use of analgesics for headache relief.
correlated her symptoms to coffee consumption. In retrospect, Hamilton Anxiety Scale scores obtained after the patient’s
she was able to trace symptom onset to the purchase of a fresh- initial psychiatric assessment were significantly elevated. He
drip coffeepot. Because this coffee was “so much better,” she proudly described his habit ofbeing “the first one in the office in
had begun consuming an average of 10 to 12 cups of strong, the morning and the last one to go at night.” When questioned
black coffee a day-more than 1,000 mg. of caffeine. about caffeine use, he responded as if it were a reflection of his
She withdrew completely from her coffee regimen and within masculinity: “I can easily put away 10 to 15 cups a day. I drink
36 hours virtually all symptoms disappeared, including her car- more coffee than anyone in my office.” He also reported that he
diovascular arrhythmias. She complained of fatigue for one often drank tea and cola. His headaches appeared at a variety
week but then began noting that she was “truly awake in the of times, averaging three or four a week. These often occurred
morning for the first time in years.” She was later “challenged” on weekends. They were generalized, constant, and throbbing
twice with caffeine after periods of abstinence. Subjective anx- and usually persisted for several hours or until relieved by anal-
iety symptoms and PVCs documented by EKG recurred. Proud gesics. A significant diagnostic clue was the analgesics he
of her diagnostic skills, she vowed to refrain from excessive use named as effective, all of which contained caffeine: I get some “

ofcoffee or tea in the future. A two-year follow-up revealed that relief from APCs, Excedrin, Vanquish, Darvon compound.
symptoms have never recurred. Fiorinal, or Cope, but regular aspirin or Darvon won’t touch
it.” He characteristically consumed 8 to 10 over-the-counter
Case 2. An ambitious 37-year-old Army lieutenant colonel headache tablets a day. Including coffee, tea, colas, and medica-
was referred from a military medical clinic to a psychiatric out- tions, his caffeine intake averaged approximately 1.500 mg. a
patient facility because of a two-year history of “chronic anx- day. The patient was provided with a list of offending in-
iety.” The symptoms, which occurred almost daily, included gredients and instructed to avoid all xanthine derivatives. After
dizziness, tremulousness, apprehension about job performance, several weeks, his headaches had almost totally disappeared
“butterflies in the stomach,” restlessness, frequent episodes of and he no longer required the use of analgesics. He later re-
“diarrhea” (two or three loose stools per day), and persistent sumed drinking coffee but rigidly limited his consumption.
difficulty in both falling and remaining asleep. Scores on the Two-year follow-up revealed almost complete alleviation of
Hamilton Anxiety Scale (21) were significantly elevated. He symptoms, including the return of anxiety scale ratings to nor-
was unable to delineate any precipitating factors, but he did mal.
note that one year previously the symptoms were accentuated
by the arrival ofa new boss who expected a 12 to 14-hour work-
day from his subordinates.
Three complete medical workups had been negative. Ten mg.
COMMENT
of chlordiazepoxide taken daily for 10 months and, later, 5 mg.
of diazepam taken daily for 4 months had produced no relief.
He disliked both medications because they “impaired his occu- The anxiety constellation of caffeinism has long been
pational precision” (primarily a desk position doing staff work). recognized. For unknown reasons, however, it has been
In reply to questioning from the psychiatrist, he described con- an object of medical concern only periodically. In Victo-
suming at least 8 to 14 cups of coffee a day (“My coffeepot is a rian England at the turn of the century, the typical coffee
permanent fixture on my desk”). He also frequently drank hot drinker was medically described as a person who was
cocoa before bedtime to relax (an average cup of cocoa contains “tremulous, loses his self-command, is subject to fits of
approximately 250 mg. of theobromine, another xanthine de- agitation and depression, and has a haggard appear-
rivative). Furthermore, his soft drink preference was exclusively
ance. . . . As with other agents, a renewed dose of the poi-
colas (3 to 4 a day). Total caffeine intake thus approximated
son gives temporary relief, but at the cost of future mis-
1,200mg. a day.
ery” (22). In America, coffee became popular during the
He responded with incredulous cynicism when informed that
caffeine toxicity might be causing his symptoms. Consequently, eighteenth century despite agitation by certain medical
he was initially unwilling (or unable) to limit his intake of cof- circles against its use (23). Medical case reports of caf-
fee, cocoa, and colas. When symptoms persisted, however, he feine toxicity (referred to as “the syndrome of coffee” by
voluntarily reduced his daily intake of caffeine, and four weeks one investigator [24]) appeared intermittently during the

Am J Psychiatr.v 131.10, October /974 1091

ANXIETY OR CAFFEINISM

early part of the twentieth century. Anxiety, vertigo, REFERENCES

headaches, nervous irritability, weak spells, and agitation
were the symptoms mentioned most frequently (24, 25). 1. Feighner JP, Robins E, Guze SB, et al: Diagnostic criteria for use
in psychiatric research. Arch Gen Psychiatry 26:57-63, 1972
One case report, however, described a psychotic reaction
2. Freedman AM, Kaplan HI, Sadock BJ (eds): Modern Synopsis of
following ingestion oflarge quantities ofcaffeine (26). Comprehensive Textbook of Psychiatry. Baltimore, Williams &
A realistic clinical perspective of caffeinism is difficult Wilkins Co, 1972, pp 330-335
to formulate. Prevalence and incidence figures are simply 3. Detre TP, Jarecki HG: Modern Psychiatric Treatment. Phila-
delphia, JB Lippincott Co. 1971, pp 219-22 1
not known. Average coffee and tea drinking patterns are
4. Chapman J: Early symptoms of schizophrenia. Br J Psychiatry
not even known. Undoubtedly, the vast majority of caf- 112:225-251, 1966
feine consumers ingest only small quantities on a daily 5. Reimann HA: Caffeinism: a cause of long-continued, low-grade fe-
basis. There is currently little evidence to suggest harm in ver. JAMA 202:1 105-1 106, 1967
these situations, and the hysteria that has intermittently 6. Ritchie JM: Central nervous system stimulants, II: the xanthines,
in The Pharmacological Basis of Therapeutics, 4th ed. Edited by
reigned throughout history seems inappropriate. Never-
Goodman LS, Gilman A. New York, Macmillan Publishing Co.
theless, all medications, including caffeine, have a poten- 1970, pp 358-370
tial for abuse, and many individuals clearly ingest symp- 7. Truitt EB: The xanthines, in Drill’s Pharmacology in Medicine, 4th
tom-producing doses daily, through use of tea, coffee, ed. Edited by Dipalma JR. New York, McGraw-Hill Book Co.
1971, pp 533-556
cola drinks, and medications. There are no certain diag-
8. Goldstein A, Kaizer S. Warren R: Psychotropic effects of caffeine
nostic clues to differentiate caffeinism from anxiety neu- in man, II: alertness, psychomotor coordination and mood. J Phar-
rosis and the two conditions could-and undoubtedly macol ExpTher l50A46-15l, 1965
do-exist simultaneously. Thus, afflicted subjects will be 9. Barr WH, Penna RP: Over-the-counter internal analgesics, in
identified only by routine collection of historical data. Handbook of Non-Prescription Drugs, 1969. Edited by Griffenha-
gen GB. Washington, DC, American Pharmaceutical Association,
Inquiry about caffeine intake during initial interviews
1969, pp 26-37
with anxious patients would probably be worthwhile, and 10. Day RL: Over-the-counter medications for menstrual problems.
including spaces for use of coffee, tea, and over-the- Ibid. pp 52-57
counter medications in operationalized mental status 1 1. Erhardt R: Psychic disturbances in caffeine intoxication. Acta Med
Scand 71:94-99, 1929
forms might identify cases as well as generate accurate
12. Oswald I: Drugs and sleep. Pharmacol Rev 20:273-303, 1968
prevalence data. 13. Colton T, Gosselin RE, Smith RP: The tolerance ofcoffee drinkers
There are several psychiatric settings in which a high to caffeine. Clin Pharmacol Ther 9:31-39, 1968
index of suspicion for caffeinism should especially pre- 14. Forrest WH, Bellville JW, Brown BW: The interaction of caffeine
vail. First, among the patients who do not seem to re- with pentobarbital as a nighttime hypnotic. Anesthesiology 36:37-
41, 1972
spond to psychopharmacological agents or nighttime
15. Goldstein A, Kaizer 5: Psychotropic effects ofcaffeine in man, III:
hypnotics, many might be simultaneously consuming a questionnaire survey of coffee drinkers and its effects in a group
large quantities of caffeine-perhaps to counteract ofhousewives. Clin Pharmacol Ther 10:477-488, 1969
drowsiness-and negating the medication’s therapeutic 16. Dreisbach RH, Pfeiffer C: Caffeine-withdrawal headache. J Lab
ClinMed28:12l2-l2l8, 1943
effects. Theoretical support for this possibility is the find-
17. Harrie JR: Caffeine and headache (letter to the editor). JAMA
ing that caffeine may induce an increased catecholamine 213:628, 1970
release (27). Other “high-risk” patients include those 18. Flynn iT: Arrhythmias related to coffee and tea (letter to the cdi-
complaining of psychophysiological symptoms (which tor). JAMA 21 1:663, 1970
caffeinism closely mimics), those hospitalized in coro- 19. Silver W: Insomnia, tachycardia and cola drinks. Pediatrics
47:635-637, 1971
nary care units (where a caffeine-induced arrhythmia
20. Ross BD: Caffeine and fluid retention (letter to the editor). JAMA
might be fatal [19]), those with recurrent headaches 218:596, 1971
(where the vicious cycle of treating caffeine withdrawal 21. Hamilton M: The assessment of anxiety states by rating. Br J Med
headaches with analgesics containing caffeine cannot be Psychol 32:50-55, 1959
altered until recognized), and children reported to be 22. Siegerist HE: Literary controversy over tea in 18th-century Eng-
land. Bull Hist Med 13:185, 1943
overly active or hyperkinetic (28). (Interestingly, caf- 23. Mueller WL: Coffee, in Encyclopedia Americana, international ed,
feine has now been reported to be therapeutic among vol VII. New York, Americana Corp. 1967, pp 209-211
accurately diagnosed hyperkinetic children [29].) 24. Orendorif 0: Letter to the editor. JAMA 62:1828, 1914
Prevalence and incidence studies of caffeinism appear 25. Powers H: The syndrome of coffee. Medical Journal and Record
121:745-749, 1925
indicated. In addition, researchers should probably con-
26. McManamy MC, Schube PG: Caffeine intoxication: report of a
sider caffeine use as a contaminating variable when eval- case the symptoms of which amounted to a psychosis. New EngI J
uating psychopharmacological or hypnotic agents. From Med 2 15:616-620, 1936
the clinical perspective, many individuals complaining of 27. Levi L: Effect of coffee on the function of the sympatho-adrenome-
anxiety will continue to receive substantial benefit from dullory system in man. Acta Med Scand 181:431-438, 1967
28. Damrau F, Damrau AM: Use of soft drinks by children and adoles-
psychopharmacological agents. For an undetermined
cents. Rocky Mt Med J 60:37-39, 1963
number of others, subtracting one drug-caffeine-may 29. Schnackenberg RC: Caffeine as a substitute for schedule II stimu-
be of greater benefit than adding another. lants in hyperkinetic children. Am J Psychiatry 130:796-798, 1973

1092 AmJ Psychiatry 131:10, October 1974