caffeine withdrawal article

Psychopharmacology (2004) 176: 1–29DOI 10.1007/s00213-004-2000-x

REVIEW

Laura M. Juliano . Roland R. Griffiths

A critical review of caffeine withdrawal: empirical validation

of symptoms and signs, incidence, severity, and associated

features

Received: 9 March 2004 / Accepted: 24 July 2004 / Published online: 21 September 2004# Springer-Verlag 2004

Abstract Rationale: Although reports of caffeine with-drawal in the medical literature date back more than170 years, the most rigorous experimental investigationsof the phenomenon have been conducted only recently.Objectives: The purpose of this paper is to provide acomprehensive review and analysis of the literatureregarding human caffeine withdrawal to empiricallyvalidate specific symptoms and signs, and to appraiseimportant features of the syndrome. Methods: A litera-ture search identified 57 experimental and 9 survey studieson caffeine withdrawal that met inclusion criteria. Themethodological features of each study were examined toassess the validity of the effects. Results: Of 49 symp-tom categories identified, the following 10 fulfilledvalidity criteria: headache, fatigue, decreased energy/activeness, decreased alertness, drowsiness, decreasedcontentedness, depressed mood, difficulty concentrating,irritability, and foggy/not clearheaded. In addition, flu-likesymptoms, nausea/vomiting, and muscle pain/stiffnesswere judged likely to represent valid symptom categories.In experimental studies, the incidence of headache was50% and the incidence of clinically significant distress orfunctional impairment was 13%. Typically, onset ofsymptoms occurred 12–24 h after abstinence, with peakintensity at 20–51 h, and for a duration of 2–9 days. Ingeneral, the incidence or severity of symptoms increasedwith increases in daily dose; abstinence from doses as lowas 100 mg/day produced symptoms. Research is reviewed

indicating that expectancies are not a prime determinant ofcaffeine withdrawal and that avoidance of withdrawalsymptoms plays a central role in habitual caffeineconsumption. Conclusions: The caffeine-withdrawalsyndrome has been well characterized and there issufficient empirical evidence to warrant inclusion ofcaffeine withdrawal as a disorder in the DSM and revisionof diagnostic criteria in the ICD.

Keywords Caffeine . Abstinence . Cessation .Deprivation . Withdrawal . Headache . Physicaldependence . DSM . ICD . Humans

Introduction

Caffeine is the most widely used behaviorally active drugin the world (Gilbert 1984). In North America, 80–90% ofadults report regular use of caffeine (Gilbert 1984; Hughesand Oliveto 1997). Mean daily intake of caffeine amongcaffeine consumers in the United States is about 280 mg,with higher intakes estimated in some European countries(Gilbert 1984; Barone and Roberts 1996). In the UnitedStates, coffee and soft drinks are the most commonsources of caffeine, with almost half of caffeine consumersingesting caffeine from multiple sources, including tea(Hughes and Oliveto 1997).

After oral ingestion, caffeine is rapidly and completelyabsorbed, with peak blood levels generally reached in 30–45 min (Denaro and Benowitz 1991; Mumford et al. 1996;Liguori et al. 1997a), and is quickly eliminated, with atypical half-life of 4–6 h (Denaro and Benowitz 1991).The primary mechanism of action of caffeine is compe-titive antagonism at A1 and A2A adenosine receptors(Fredholm et al. 1999). Caffeine produces a variety ofphysiological effects, including effects on the cerebralvascular system, blood pressure, respiratory functioning,gastric and colonic activity, urine volume, and exerciseperformance (James 1997). Low to moderate doses ofcaffeine (20–200 mg) produce reports of increased well-being, happiness, energy, alertness, and sociability,

L. M. JulianoDepartment of Psychology, American University,4400 Massachusetts Avenue,Washington, DC, 20016, USA

R. R. Griffiths (*)Department of Psychiatry and Behavioral Sciences, Departmentof Neuroscience, Johns Hopkins University School ofMedicine,5510 Nathan Shock Drive,Baltimore, MD, 21224-6823, USAe-mail: [email protected].: +1-410-5500034Fax: +1-410-5500030

whereas higher doses are more likely to produce reports ofanxiety, jitteriness, and upset stomach (Griffiths et al.2003). Chronic administration of caffeine results intolerance to a number of its physiological, subjective,and behavioral effects (Griffiths and Mumford 1996).Caffeine has been shown to function as a reinforcer inhumans (e.g. Hughes et al. 1991; Evans et al. 1994), andsome individuals become clinically dependent on caffeineas indicated by being unable to quit and continuing usedespite having medical problems made worse by caffeine(Strain et al. 1994; Hughes et al. 1998).

Regular use of caffeine also produces physical depen-dence, evidenced as time-limited withdrawal symptomsupon the termination or reduction of one’s usual caffeinedose. Physical dependence on caffeine has been docu-mented in both pre-clinical and clinical research, and thebiological basis has been hypothesized to be increasedfunctional sensitivity to endogenous adenosine (Griffithsand Mumford 1996). Symptoms of caffeine withdrawalhave been described in the medical literature for more than170 years. In 1988, the first comprehensive review ofclinical reports and experimental studies on caffeinewithdrawal was published (Griffiths and Woodson1988), which provided evidence for caffeine withdrawalas a discrete clinical syndrome. Since that time, theresearch literature on caffeine withdrawal has increasedsubstantially. For example, of 48 blind caffeine-withdraw-al studies identified for this review, only five werepublished before 1988.

The present review was inspired by this emergentresearch literature, which has not been comprehensivelyreviewed, as well as by the practical need to developempirically based diagnostic criteria for caffeine with-drawal. In 1994, a tentative research diagnosis of caffeinewithdrawal was proposed in the Diagnostic and StatisticalManual of Mental Disorders (DSM-IV) to encourageresearch on diagnostic criteria and the utility of thediagnosis (American Psychiatric Association 1994). Onlyone-quarter of the blind studies identified for this reviewwere available to the DSM-IV Work Group (Hughes1994). The review also addresses the previous suggestions(Rubin and Smith 1999; Dews et al. 2002) that caffeinewithdrawal is not clinically significant and is primarilydetermined by expectancies.

Methods for searching and categorizing the literature

Search strategy and inclusion criteria

The following strategies were used to identify possiblestudies for this review: (1) PubMed (1950–2004) andPsycInfo (1872–2004) searches were conducted with thekeywords “caffeine” in conjunction with “withdrawal,”“dependence,” “deprivation,” or “abstinence”; (2) theauthors searched their personal journal article collectionson caffeine withdrawal; and (3) relevant references cited inpapers obtained through the first two strategies and majorreviews of caffeine and caffeine withdrawal (Griffiths and

Woodson 1988; Fredholm et al. 1999; Nehlig 1999;Griffiths et al. 2003) were examined to identify additionalrelevant papers. To be included, experimental studies hadto require a caffeine abstinence period of 12 h or greaterbecause withdrawal onset generally occurs after 12 h ofabstinence (Griffiths and Woodson 1988).

Excluded studies

Although possibly relevant to caffeine withdrawal, severaltypes of studies were excluded from detailed summariza-tion and analysis (Tables 1, 2, 3, 4): (1) several studieswere excluded because they were not clearly interpretable(Horst et al. 1934; Mackenzie et al. 1981; Ammon et al.1983; Smith 1996; Reeves et al. 1999; Watson et al. 2000);(2) a series of carefully conducted studies comparingplacebo and caffeine conditions after overnight abstinencewere not presented in a way that was clearly interpretableas caffeine-withdrawal effects (Yeomans et al. 2000a,b,2001); (3) several studies that purported to demonstratecaffeine effects after a period of abstinence (e.g., Lieber-man et al. 1987; Brice and Smith 2002) were excluded,although some of these studies acknowledged that theeffects may represent a reversal of caffeine withdrawal(e.g., Bruce et al. 1986; Hindmarch et al. 1998; Kenemanset al. 1999; Smit and Rogers 2000); (4) studies exploringcaffeine-withdrawal headache following surgical anesthe-sia were excluded because of the possible confoundingeffects of anesthesia (Galletly et al. 1989; Fennelly et al.1991; Weber et al. 1993; Nikolajsen et al. 1994; Hampl etal. 1995); (5) case studies of neonatal caffeine withdrawalwere excluded (McGowan et al. 1988; Thomas 1988); and(6) a survey study purporting to assess the prevalence ofcaffeine-withdrawal headache was excluded because themethods were ambiguous and it was not clear whether thesubjects ever experienced periods of caffeine abstinence(Sjaastad and Bakketeig 2004).

Categorization of withdrawal studies

The inclusion strategy resulted in the identification of 42double-blind experiments (Table 1), 15 non-blind andsingle-blind experiments (Table 2), and 9 survey studies(Table 3). Table 4 provides a summary of the individualwithdrawal symptoms that are documented in Tables 1, 2,and 3 (for completeness, Appendix A shows which of theexperimental studies in Tables 1 and 2 failed to documenteach of the evaluated symptoms and signs). In construct-ing Table 4, all symptoms documented in Tables 1, 2, and3 were first catalogued, and then phenomenologicallysimilar descriptors were combined (e.g., drowsiness,sleepy, drowsy/sleepy/tired, sedated, feel half awake, anddecreased wakefulness were combined in a single categorycalled “drowsiness/sleepiness”). Categories and descrip-tors for each category are presented below. It is recognizedthat the resulting 49 symptom categories may not representfully independent constructs. In the absence of empirical

2

Tab

le1

Sum

maryof

doub

le-blin

dexperimentalstud

iesof

caffeine

with

draw

ala

Reference

Design

With

draw

alsig

nsandsymptom

sb,c

1.Goldstein

(1964)

[experim

entscandd]

N=approxim

ately16

heavycoffe

edrinkers(5

ormorecups/day),25

moderatedrinkers(2–4

cups/day),and37

light

drinkers

(0–1

cups/day);with

in-subjectsd

esign;

subjectsabsta

ined

from

coffe

eafterlunch

andreceived

150-mgcaffe

ineor

placeboin

decaffe

inated

coffe

eat

bedtim

eover

four

consecutivenights,

with

both

treatmentsgiventwice;

abstinencenotb

iologically

verified

Inheavycoffe

edrinkers

morning

headache

occurre

dsig

nific

antly

morefre

quently

afterp

lacebo

(25%

oftrials)

than

after

caffe

ine(3%

oftrials);inmoderatedrinkers

thefre

quency

ofheadache

was

non-sig

nific

antly

higher

afterp

lacebo

(12%

oftrials)

than

afterc

affeine(7%

oftrials);inlig

htdrinkersthefre

quency

ofheadache

was

<1%

afterb

othcaffe

ineandplacebo

2.Goldstein

etal.(1969)

N=38

femaledaily

coffe

edrinkers

(5or

morecups/day)a

nd18

femalecoffe

eabsta

iners;with

insubjectsandacross

group

desig

n;subjectsabsta

ined

from

caffe

ineafterd

innera

ndreceived

placebo,150-mgor

300-mgcaffe

inein

decaffe

inated

coffe

ethefollo

wingmorning

(about

0900

hours);treatmentswererepeated

threetim

es;a

bstin

ence

notb

iologically

verified

Comparedwith

coffe

eabsta

iners,coffe

euserswereless

alert,activ

e/energetic,a

ndcontent,andmoresle

epy,

irritable,

and

nervousa

fterc

affeineabstinence;am

ongcoffe

eusersplaceboproduced

increasedratin

gsof

headache,lazy/slu

ggish

,irritable,

sleepy,

anddecreasedratin

gsof

talkative,

contentedness,andenergy/activecomparedwith

caffe

ine;

caffe

inegenerally

produced

adose-re

latedsuppressionof

with

draw

alsymptom

sin

thecoffe

eusers

3.Ro

bertson

etal.(1981)

N=18

adultswho

werenoth

abitu

alcaffe

inec

onsumersw

erecaffe

ineabstinent

for3

weeks

precedingthes

tudy;w

ithin

subjects

andacross

groupdesig

n;approxim

atelyhalfthesubjectsreceived

750mgcaffe

ine/dayin

flavoreddrinks

for7

days,followed

bysubstitutionof

placebofla

voreddrinks

for4

days;a

bstin

ence

was

biologically

verified

Substitutionof

placebodidnotresultindetectable

effectson

bloodpressure,u

rinarynorepinephrin

e,or

urinaryepinephrine

4.Griffiths

etal.(1986)

N=7males

who

wereheavycoffe

edrinkers

(mean12

cups/day);with

in-subjectsdesig

n;aftera

pproximately10

consecutive

days

ofad

libitu

mcoffe

econsum

ption(m

ean1.25-g/day

caffe

ineforthe

last5days),subjectsweresw

itchedto

decaffe

inated

coffe

efora

tleast10

days;a

bstin

ence

notb

iologically

verifiedbutstudy

was

conductedin

aresid

entiallaboratory

Comparedwith

caffe

inated

coffe

econsum

ption,

decaffe

inated

coffe

esubstitutionresultedin

increasedsubjectratings

ofheadache,sleepy,lazy,fatigue

(POMS)

anddecreasedalert,activ

e,vigor(PO

MS),and

friendliness(POMS);o

bservatio

nof

subjects’

behavior

also

show

edsig

nific

antw

ithdraw

albasedon

acompositewith

draw

alscore;

cigarette

smokingdecreased

durin

gdecaffe

inated

coffe

econsum

ption;

100%

ofsubjectsreporte

dwith

draw

alheadache

5.Griffiths

etal.(1990)[Ph

ase1]

N=7investigator-s

ubjectswith

histo

riesof

regularc

affeineconsum

ption;

with

in-subjectsdesig

n;subjectsgivencapsules

containing

100mgcaffe

ine/dayfor9

–14days,thenplacebocapsules

for1

2days,a

ndthen

capsules

containing

100mg

caffe

ine/dayfor7

–12days;a

bstin

ence

was

biologically

verified

Comparedwith

caffe

inemaintenance,p

lacebo

substitutionresultedin

increasedheadache,c

erebralfullness,irritable/cross/

grum

py,d

epression,

musclepain/stiffness,lethargy/fatig

ue/tired/slu

ggish

,craving,and

flu-like

feelings,a

nddecreasedalert/

attentive/observant,well-b

eing,sociald

isposition,m

otivationforw

ork,

concentra

tion,

energy/active,

urge

todo

tasks/w

ork,

content/satisfied,

andself-confidence;these

with

draw

alsymptom

speakedon

days

1or

2andprogressivelydecreasedtoward

pre-with

draw

allevelsover

abouta

week

6.Griffiths

etal.(1990)[Ph

ase2]

N=7investigator-s

ubjectswith

histo

riesof

regularc

affeineconsum

ption;

with

in-subjectsdesig

n;subjectsgivencapsules

containing

100mgcaffe

ine/dayfor6

weeks;o

nfiv

eoccasio

nsplacebocapsules

weresubstituted

forc

affeinefor1

day

(separated

byan

averageof

9days);abstinencewas

biologically

verified

Comparedwith

caffe

inedays,intermittentp

lacebo

resultedin

increasedfatig

ue(POMS)

d ,confusion–bewild

erment(PO

MS),

andtotalm

ooddistu

rbance

(POMS),a

nddecreasedvigor(PO

MS)

andfriendliness(POMS);inadditio

n,comparedwith

caffe

inedays,intermittentp

lacebo

resultedin

allo

fthe

with

draw

alsymptom

sobserved

inphase1(see

above)

aswella

sincreasedmuzzy,d

rowsy/sleepy,y

awning,b

lurre

dvisio

n,lig

htheaded/dizzy,impaire

dwork/thoughtrelated

activ

ities

and

impaire

dverbal

ability;a

llof

thesewith

draw

alsymptom

splus

decreasedanxious/n

ervous

werealso

signific

antw

hen

interm

ittentp

lacebo

was

comparedwith

achronicplaceboperio

d(days8–

12of

placebofro

mphase1)

[exceptio

ns:

lightheaded/dizzy,impaire

dverbal

ability,u

rgeto

dotasks/w

orkandmotivationforw

orkwerenotsignific

ant];

with

in-

subjectsanalysisof

both

interm

ittentp

lacebo

versus

caffe

ineandof

interm

ittentp

lacebo

versus

chronicplacebodemonstrated

thatmorethan

57%

ofsubjectsshow

edincreasedheadache,m

uzzy,d

rowsy/sleepy,impaire

dwork/thoughtrelated

activ

ities,

anddecreasedsocial

disposition,c

oncentratio

n,andself-confidence.

7.vanDusseldorpandKatan

(1990)

N=45

daily

coffe

edrinkers

(4–6

cups/day);with

in-subjectsdesig

n;subjectsreceived

5-cups/day

caffe

inated

coffe

e(435

mg

caffe

ine/day)

ordecaffe

inated

coffe

e,each

for6

weeks;a

bstin

ence

notb

iologically

verified

Headachewas

signific

antly

morelik

elyto

occurd

uringthefirstweekof

thedecaffe

inated

coffe

eperio

d;42%

ofsubjects

reporte

dheadache;h

eadaches

starte

don

the1sto

r2nd

dayof

abstinenceandlaste

d1–6days

(mean2.3days)

8.Bruceet

al.(1991)

N=21

habitual

caffe

ineusers(m

ean389mg/day);w

ithin

subjectsandacross

groupdesig

n;subjectsin

onegroupabsta

ined

from

caffe

inefor2

4-hpriortothreesessions

conductedatweeklyintervals;subjectsin

anotherg

roup

absta

ined

from

caffe

ine

for7

days

before

aswella

sbetweenthreesessions

conductedat

48-h

intervals;subjectsin

both

groups

received

capsules

containing

placebo,

250-mgor

500-mgcaffe

inein

random

order;abstinencewas

biologically

verified

Afte

rplacebo,subjectsinthe24-h

abstinencegroupreporte

dincreasedheadache

andtiredness

andshow

edmorepsychomotor

perfo

rmance

tiringthan

the7-dayabstinencegroup;

inthe24-h

abstinencegroup,

comparedwith

either

dose

ofcaffe

ine,

placeboincreasedheadache

andtiredness

onunipolar

scales,a

ndincreasedlethargy

anddrow

sinesson

bipolars

cales

(lethargic–energetic;d

rowsy–alert)

9.Silverman

etal.(1992)

N=62

daily

caffe

ineconsum

ers(m

ean235mg/day);w

ithin-subjectsdesig

n;subjectswereassessed

while

consum

ingtheir

norm

alcaffe

ineintake

(baseline)

andaftertwo2-dayperio

dsdurin

gwhich

they

consum

edcaffe

ine-fre

edietsandwere

administered

capsules

containing

caffe

ine(m

atched

totheiru

sual

intake)o

rplacebo;a

bstin

ence

was

biologically

verified

Comparedwith

both

baselin

eandcaffe

ine,placeboincreasedirritable/cross/g

rumpy,b

lurre

dvisio

n,drow

sy/sleepy,y

awning,

lethargy/fa

tigued/tired/sluggish,m

uzzy/fo

ggy/notclearheaded,headache,flu

-like

feelings,heavy

feelingin

armsa

ndlegs,hot

andcold

spells,

Beck

DepressionInventoryscores,som

atization(SCL

-90-R),state

andtra

itscales

ofState-TraitA

nxiety

Inventory,

fatig

ue(POMS),c

onfusio

n–bewild

erment(PO

MS),total

mooddistu

rbance

(POMS),a

nddecreasedwell-b

eing,

desireto

socialize,

talkativeness,urge

todo

tasks/w

ork-relatedactiv

ities,e

nergy/activ

e,content/satisfied,

self-confidence,

vigor(PO

MS),frie

ndlin

ess(PO

MS),p

sychom

otor

perfo

rmance

(tappingspeed);com

paredwith

baselin

eandcaffe

ine,durin

gtheplaceboperio

dmoresubjectsreporte

dmoderateor

severe

headache

(52%

),used

analgesic

sdespite

discouragement

(13%

),andshow

edabnorm

alscores

ontra

it-anxiety(8%),vigor(11%),andfatig

ue(8%),andtheBe

ckDepressionInventory

(11%

)10.E

vans

andGriffiths

(1992)

N=32

daily

caffe

ineconsum

ers(m

ean343mg/day);w

ithin

subjectsandacross

groupdesig

n;subjectsreceived

capsules

throughout

thestu

dy;subjectsin

thechronicplacebocondition

(N=1

6)received

placebofor1

8days,followed

bya3-day

caffe

ineversus

placebochoice

phase,

follo

wed

byanother7

days

ofplacebo;

subjectsin

thechroniccaffe

inecondition

(N=1

6)received

caffe

inefor1

8days

(thedose

progressed

from

100-mgt.i.d.to300-mgt.i.d.);

follo

wed

bya3-daycaffe

ine

versus

placebochoice

phase;follo

wed

by2days

ofprogressivelylower

doseso

fcaffeine(approximately183-mgt.i.d.o

nthe

2ndday),followed

by5days

ofplacebo;

abstinencewas

biologically

verified

Comparedwith

subjectsin

thechronicplacebocondition,subjectsinthechroniccaffe

inecondition

reporte

dhigher

ratin

gsof

headache

andlower

ratin

gsof

talkativenessin

thefin

alwith

draw

alperio

d;with

inthecaffe

inegroup,

noeffectsoccurre

dbetweentheplaceboandcaffe

inechallengedays

which

was

interpretedas

indicatin

gthat24

hof

abstinencewas

insufficient

toproducewith

draw

alsymptom

supon

term

inationof

very

high

caffe

inedoses(i.e.,9

00mg/day)

3

Reference

Design

With

draw

alsig

nsandsymptom

sb,c

11.H

ughese

tal.(1993)

[sum

marizes

results

from

Hughese

tal.(1991,1992);Oliv

etoet

al.(1992a,b)]

N=37

daily

coffe

edrinkers

(mean552mg/day)

who

reporte

dusingcoffe

eforc

affeineeffectsor

show

edwith

draw

alor

preference

forc

affeinein

ascreeningtest;

with

in-subjectsd

esign;

sixtrialsa

tweeklyintervals;on

oned

aysubjectsconsum

eddecaffe

inated

coffe

eandon

anotherd

ayconsum

edcaffe

inated

coffe

e(100

mg/serving);assessm

ento

ccurredaftera

bout

24-h

abstinence;

abstinencenotb

iologically

verifiedin

allsubjects

With

in-subjectssta

tistical

analyses

show

edthat

30%,2

4%,2

4%,a

nd49%

ofsubjects,

respectiv

ely,

reporte

dsig

nific

ant

increasesin

headache,d

rowsin

ess,fatig

ue,o

rone

ormoreof

thesesymptom

son

decaffe

inated

days

comparedwith

caffe

inated

days

12.S

train

etal.(1994)

N=11

daily

caffe

ineusers(m

edian357mg/day)

who

met

DSM

-IVdiagnosticcriteria

fors

ubsta

ncedependence

oncaffe

ine;

with

in-subjectsdesig


aftertwo2-dayperio

dsdurin

gwhich

they

consum

edcaffe

ine-fre

edietsand

wereadministered

capsules

containing

caffe

ine(m

atched

totheiru

sual

intake)o

rplacebo;a

bstin

ence

was

biologically

verified

Overall,

82%

show

edevidence

ofcaffe

inewith

draw

aldurin

gplacebo;

64%

reporte

dmaxim

alratin

gsof

headache;c

ompared

with

standardizednorm

s,27%

show

edextre

meincreasesin

fatig

ue(POMS),4

5%show

edextre

medecreasesin

vigor

(POMS),and

36%

show

edelevations

indepressio

n(BeckDepressionInventory);com

paredwith

caffe

ine,afterp

lacebo

55%

ofsubjectsshow

edsig

nific

antd

ecreases

inpsychomotor

perfo

rmance

(tappingspeed);4

5%reporte

danalgesic

usedespite

discouragement;73%

reporte

dfunctio

nalimpairm

entinnorm

aldaily

activ

ities

13.B

rauere

tal.(1994)

eN=

11daily

coffe

econsum

ers(mean480mg/day);w

ithin-subjectsd

esign;

over

a33-h

perio

dsubjectsconsum

edacaffe

ine-fre

ediet

andwereadministered

either

capsules

containing

placeboor

300-mgcaffe

ine;

abstinencewas

biologically

verified

Comparedwith

caffe

inemaintenance,3

3-habstinenceincreasedheadache

14.H

aleet

al.(1995)

N=18

adolescents(ages

11–15),d

aily

coladrinkers(m

ean2.3,12-ozcaffe

inated

colas/d

ay);with

in-subjectsd

esign;

sixtrialsa

tweeklyintervals;on

onedaysubjectsconsum

edcaffe

ine-fre

ecolasa

dlib

itum

andon

anotherd

ayconsum

edcaffe

inated

colas

adlib

itum

(33mg/8oz);assessmento

ccurredaftera

bout

24-h

abstinence;

abstinencenotb

iologically

verified

Comparedwith

thecaffe

ine,placeboincreaseddepressed,

drow

sy–sleepy,andfatig

ue–tire

d,anddecreasedirregular

heartbeat

andmoretalkative;

with

in-subjectsanalyses

revealed

that

33%

ofsubjectsshow

edsig

nific

ante

ffectson

oneor

more

symptom

ratin

gs15.H

ugheset

al.(1995)

N=11

daily

coffe

edrinkers(m

ean5.6cups/day)selectedforshowingcaffe

ineabstinencesymptom

sorc

affeinepreference

ina

screeningtest;

with

in-subjectsdesig

n;six

trialsat

weeklyintervals;on

onedaysubjectsconsum

eddecaffe

inated

coffe

ead

libitu

mandon

anotherd

ayconsum

edcaffe

inated

coffe

eadlib

itum

(eith

er25

mgor

50mg/serving);assessm

ento

ccurredafter

about2

4-habstinence;

abstinencenotb

iologically

verifiedf

Comparedwith

thecaffe

ine,

placeboincreaseddrow

sy–sleepy–tired,h

eadaches,fatigue–tire

d,lazy–sluggish

,and

depressed

anddecreasedstimulated–active–energetic–excited,

alert–attentive–able

toconcentra

te,a

ndvigor;with

in-subjectsanalyses

revealed

that

73%

ofsubjectsshow

edsig

nific

ante

ffectson

oneor

moresymptom

ratin

gs,w

ith45%

reporting

increasesin

drow

sy–sleepy–tired;2

7%reporting

decreasesin

stimulated–active–energetic–excited,

alert–attentive–able

toconcentra

te,

andvigor;18%

reporting

increasedfatig

ued–

tired,d

epressed,a

nddecreasedcontent–relaxed–

satisfie

d,and9%

reporting

increasedheadache,irritable–fru

strated–angry–cross,lazy–

sluggish

,and

decreasedtalkative

16.M

itchellet

al.(1995)

N=9daily

coffe

econsum

ers(m

ean686mg/day);w

ithin-subjectsdesig

n;subjectsconsum

edacaffe

ine-fre

ediet

andwere

administered

capsules

containing

placeboor

caffe

ine(50%

or100%

oftheiru

sualam

ounts)on

thedaybefore

andthedayof

assessment;abstinencewas

biologically

verified

Comparedwith

the50%

and100%

caffe

ineconditions,placebo(33.5-habstinence)

increasedheadachy,sluggish

,and

tired–

sleepy;

abstinencedidnota

ffect

amount

ofresponding

onan

operanttaskreinforced

bycoffe

edeliv

ery

17.R

ichardsonet

al.(1995)

N=49

regularc

affeineconsum

ers(m

ean252mg/day)

andN=

18non-consum

ers;four

groups:n

on-consumersandthree

consum

ergroups

who

absta

ined

from

caffe

inefore

ither

90min,o

vernight,o

r7days;subjectsin

each

groupreceived

capsules

containing

placebo,

70mgor

250mgin

counterbalancedorder;abstinencenotb

iologically

verifiedf

Overnight

abstinence(13–15

h)comparedwith

90-m

inabstinenceproduced

increasedratin

gsof

angry,tired,d

rowsy,d

ejected

anddecreasedratin

gsof

clearheaded,

friendly,

andcheerfu

l;overnighta

bstin

ence

comparedwith

7days

ofabstinence

produced

increasedratin

gsof

angry,

tired,d

rowsy

anddejected;o

vernight

abstinencecomparedwith

non-consum

ers

produced

increasedheadache,a

ngry,tire

d,drow

sy,d

ejectedanddecreasedclearheaded

18.R

ogerset

al.(1995)

N=24

low

caffe

ineusers(mean47

mg/day)

andN=

25moderatecaffe

ineusers(mean205mg/day);ten

experim

entalsessio

ns;

aftero

vernight

abstinencehalfthesubjectsconsum

edadrinkcontaining

70-m

gcaffe

ineandhalfconsum

edanon-caffe

inated

drink;

assessmento

ccurred1hlater;abstinencenotb

iologically

verified

Inmoderatecaffe

ineusers,comparedwith

subjectsreceivingcaffe

ine,

subjectsreceivingplaceboreporte

dincreasesin

tired

anddecreasesin

cheerfu

l,clearheaded,

energetic,a

ndliv

ely;

incontrast,

nosig

nific

antd

ifferenceswerefoundbetween

caffe

ineandplaceboin

low

caffe

ineusers

19.S

treuferte

tal.(1995)

N=25

caffe

ineuserswho

held

managerialp

ositions(m

ean565mg/day);w

ithin-subjectsdesig

n;subjectswereinstructedto

absta

infro

mcaffe

inefor3

6-hpriortoeach

of2assessmentd

ays;thedaypriortotheassessmentd

ayandmorning

ofthe

assessmentd

aysubjectsweregiveneither

capsules

containing

caffe

ine(m

atched

totheiru

sualintake)o

rplacebo;abstin

ence

was

biologically

verified

Comparedwith

caffe

ine,

placebowas

associated

with

increaseddrow

sy/sleepy,y

awning,lethargy/fatig

ue/tired/slu

ggish

,irritable/cross/g

rumpy

anddecreasedalert/attentive/observant,well-b

eing,a

bilityto

concentra

te,e

nergy/activ

e,need

topass

water

frequently

g ;comparedwith

caffe

ineplacebodecreasedsysto

licbloodpressure

anddecreasedperfo

rmance

oncomplex

simulated

tasks(i.e.,n

umbero

fdecisions,speedof

response

toinform

ation,

diversity

ofactio

n,numbero

fbackw

ard

integrations,a

ndappliedinitiative)

20.L

ader

etal.(1996)h

N=40

habitual

caffe

ineconsum

ers(m

ean360mg/day);w

ithin

andacross

groups

staggered

cohortdesig

n;afterb

aseline

assessmentd

uringusualc

affeineconsum

ption,

subjectsreceived

capsules

containing

either

placeboor

caffe

ine(m

atched

totheiru

sual

intake);subjectsreceived

placebocapsules

for2

consecutivedays

follo

wed

bycaffe

inecapsules

for2

days

or3

days;a

bstin

ence

was

biologically

verified

Comparedwith

baselin

e,placeboincreasedtiredness,lethargy,

sedatio

n,andmuzziness,a

nddecreasedalertnessand

concentra

tion;

theseeffectswerereversed

upon

caffe

inere-adm

inistratio

n;comparedwith

baselin

e,placeboincreasedEE

Galphapeak

magnitude

andratin

gsof

sleep

quality,a

nddecreasedskin

conductance,

systo

licbloodpressure,a

ndratin

gsof

onsettosle

ep;6

8%of

subjectsreporte

dincreasedtired

andlethargy;4

5%reporte

ddiffu

sethrobbingheadache,w

ith28%

ofthesealso

reporting

nausea

andsic

kness

21.C

omer

etal.(1997)

N=12

daily

caffe

ineusers;with

in-subjectsd

esign;

17-day

inpatient

study;subjectsg

iven

capsules

containing

300-mgcaffe

ine

daily

except

fortwo2-dayplacebosubstitutions

ondays

5–6and12–1

3;abstinencenotb

iologically

verifiedbutstudy

was

conductedin

aresid

entiallaboratory

Onplacebodays

subjectsreporte

dincreasedratin

gsof

headache,m

iserable,sedated,

sleepy,tired,u

nmotivated,and

yawning,

anddecreasedratin

gsof

alert,anxious,energetic,a

ndself-confident

22.L

iguoriet

al.(1997b)

[Exp

1]N=

8daily

caffe

inated

cola

consum

ers(m

ean157mg/day);w

ithin-subjectsdesig

n;six

trialsat

weeklyintervals;on

oneday

subjectsconsum

edcaffe

ine-fre

ecolasa

dlib

itum

andon

anotherd

ayconsum

edcaffe

inated

colasa

dlib

itum

(eith

er33

mgor

17mg/8oz);assessmento

ccurredaftera

bout

24-h

abstinence;

abstinencenotb

iologically

verifiedf

Comparedwith

the33-m

gcondition,p

lacebo

decreasedactiv

e/energy/excited,

moretalkative,

motivated

towork,

andwell-

being;

comparedwith

the17-m

gcondition,p

lacebo

increasedanxious/tense/nervous,a

nxiety

(POMS),c

onfusio

n(POMS),

irritable/fru

strated/angry/cross,a

ndsto

machache/upsetstomach,

anddecreasedwell-b

eing,frie

ndlin

ess(POMS),a

ndvigor

(POMS);w

ithin-subjectanalysisrevealed

that50%

ofthesubjectsin

the33-m

gcondition

and25%

ofthesubjectsin

the17-

mgcondition

show

edheadache,d

rowsin

ess,or

fatig

ue(63%

show

eddrow

sinessand/or

fatig

uein

oneor

both

ofthe

conditions)

Tab

le1(con

tinued)

4

Reference

Design

With

draw

alsig

nsandsymptom

sb,c

23.L

iguoriet

al.(1997b)

[Exp

2]N=

16daily

coffe

eandcola

consum

ers(m

ean579mg/day);w

ithin-subjectsdesig

n;six

trialsat

weeklyintervals;on

oneday

subjectsconsum

edcaffe

ine-fre

ecolasad

libitu

mandanotherd

ayconsum

edcaffe

inated

colasad

libitu

m(33mg/8oz);

assessmento

ccurredaftera

bout

24-h

abstinence;

abstinencenotb

iologically

verifiedf

Comparedwith

thecaffe

ine,

placeboincreaseddrow

sy/sleepy,fatigue/tired,

headache,lazy/slu

ggish

/slow

-moving,

and

stomachache/upsetstomach,

anddecreasedactiv

e/energetic/excited,

confident,motivated

towork,

well-b

eing,a

ndperfo

rmance

onapsychomotor/cognitiv

etask;d

rowsin

essandheadache

ordrow

sinessandfatig

uereliablyoccurre

din

2of

16participants

24.L

iguoriandHughes(1997)

N=11

daily

coffe

eandcolaconsum


ithin-subjectsd

esign;

subjectswereteste

din

four

caffe

ineversus

nocaffe

ineconditionsusingprocedures

similartoLiguorie

tal.(1997b);thefour

caffe

ineconditionswerecola

33mg/

serving,

coffe

e33

mg/serving,

cola100mg/serving,

coffe

e100mg/serving;

assessmento

ccurredaftera

bout

24-h

abstinence;

abstinencewas

biologically

verified

Comparedwith

thecaffe

ine,placeboincreaseddrow

sy/sleepy,fatigue/tired,

lazy/sluggish/slow

-moving,

andnausea/vom

iting,

anddecreasedactiv

e/energetic/excited,

alert/attentive/able

toconcentra

te,c

onfid

ent,fre

quentu

rinationg,impatience,

motivated

towork,

stronger/m

orevigorous/m

oreenergy,w

ell-b

eing,a

ndperfo

rmance

onapsychomotor/cognitiv

etask

25.S

chuh

andGriffiths

(1997)

N=20

regularc

affeineconsum

ers(m

ean379mg/day);w

ithin-subjectsdesig


aftertwo19-h

perio

dsdurin

gwhich

they

consum

edcaffe

ine-fre

edietsandwereadministered

capsules

containing

caffe

ine(m

atched

totheiru

sual

intake)o

rplacebo;a

bstin

ence

was

biologically

verified

Comparedwith

caffe

ine,placeboincreasedheadache,w

orn-out,andflu

-like

feelings,and

decreasedalert,well-b

eing,h

elpful,

andupsetstomach;

onadrug

versus

money

choice

procedure,subjectschoseto

forfe

itmoney

toavoidreceivingplacebo—

thiswas

signific

antly

diffe

rent

from

thecaffe

inecondition

26.G

arrettandGriffiths

(1998)

N=28

adults;

with

in-subjectsd

esign;

each

subjectw

asexposedto

four

conditionse

achof

which

involved

a9–12

dayexposure

(capsulescontaining

300mg/70

kgcaffe

ineor

placebo)

follo

wed

bya2-daychallenge(caffeineor

placebo):(1)

acute

abstinence(caffeinefollo

wed

byplacebo),(2)

chronicabstinence(placebo

follo

wed

byplacebo),(3)

acutecaffe

ine(placebo

follo

wed

bycaffe

ine),(4)

chroniccaffe

ine(caffeinefollo

wed

bycaffe

ine);a

bstin

ence

was

biologically

verified

Placebochallengeafterc

affeine(acute

abstinence)

resultedin

increasedconfusion–bewild

erment(PO

MS),fatigue

(POMS),

totalm

ooddistu

rbance

(POMS)

anddecreasedvigor(PO

MS)

comparedwith

each

oftheotherthree

conditions;acute

abstinenceresultedin

decreasedalert/attentive/observant,well-b

eing,abilityto

concentra

te,and

energy/activecomparedwith

thecaffe

inechallengeafterc

hronic

abstinenceandchroniccaffe

ine,

andincreasedirritable/cross/g

rumpy,d

epressed,a

nddrow

sy/sleepy

relativ

eto

placebochallengeafterc

hronic

abstinence;

subjectswerewillingto

forfe

itmoremoney

toavoid

receivingplacebowhenmaintainedon

caffe

ine(acute

abstinence)

comparedwith

chroniccaffe

ineandchronicplacebo

27.Jam

es(1998)

N=36

habitual

caffe

ineconsum

ers(about

370mg/day);w

ithin-subjectsdesig

n;subjectsreceived

capsules

throughout

the

study;fourc

onditio

ns:6

days

ofcaffe

inefollo

wed

by1dayof

placebo(acute

abstinence),6

days

ofcaffe

inefollo

wed

by1

dayof

caffe

ine(habitu

aluse),6

days

ofplacebofollo

wed

by1dayof

caffe

ine,

6days

ofplacebofollo

wed

by1dayof

placebo(chronic

abstinence);a

bstin

ence

was

biologically

verified

Subjectsin

theacuteabstinencecondition

reporte

dmorefre

quenth

eadaches

(47%

)thansubjectsin

thehabitualusecondition

(14%

)and

inthec

hronicabstinencec

onditio

n(31%

);subjectsin

acutea

bstin

ence

condition

also

experie

nced

mores

evereand

longer

headaches,less

alertness,bettersleep

quality

andduratio

n,andworsenedperfo

rmance

onacharacterrecognitio

ntask

than

theaverageof

thethreeotherc

onditio

ns28.P

hillips-ButeandLa

ne(1998)

N=31

daily

coffe

econsum


ithin-subjectsd

esign;

subjectsabsta

ined

from

caffe

ineovernightand

then

consum

edeither

capsules

containing

250-mgcaffe

ineor

placebo4-hpriortoassessment;thus

4-habstinencewas

compared

with

anestim

ated

12–28habstinence;

abstinencenotb

iologically

verified

Comparedwith

4-habstinence,

overnighta

bstin

ence

increasedfatig

ue(POMS),sleepy,

andyawning,a

nddecreasedvigor

(POMS)

andsysto

licanddiastolic

bloodpressure

29.R

obelin

andRo

gers

(1998)

N=64

daily

caffe

ineconsum

ers(mean453mg/day);b

etweengroups

desig

n;aftero

vernight

abstinence,four

groups

ofsubjects

consum

edthreefru

it-fla

voreddrinks

spaced

at75–9

0min

intervals;conditionsw

erethreeplacebodrinks

or1,

2,or

3drinks

containing

caffe

ine(1.2-m

g/kg

orabout8

6-mgcaffe

ine);finalassessmento

ccurredafter≥

17hof

abstinence;abstinencenot

biologically

verifiedf

Comparedwith

subjectswho

received

1,2,

or3drinks

ofcaffe

ine,subjectswho

received

placeboshow

eddecreasedenergetic

moodandslo

wer

reactio

ntim

eperfo

rmance

30.V

anSo

eren

andGraham

(1998)

N=6maleathletes

who

weredaily

caffe

ineconsum

ers(m

ean761mg/day);w

ithin-subjectsdesig

n;subjectswereteste

dafter

receivingcapsules

containing

caffe

ineor

placeboafter0

,2,o

r4days

ofabstinencefro

mdietarycaffe

ine;

sixtrialswere

separatedby

≥10days;a

bstin

ence

was

biologically

verified

Alth

ough

notstatistically

analyzed,the

authorsreportedthatallsubjectsreportedwith

draw

alsymptom

slastin

gfro

m2days

to4days

includingsevere

headaches,fatig

ue,lethargy,

andflu

-like

symptom

s;tim

eto

exhaustio

non

acyclingtask

was

not

affected

bytheduratio

nof

dietarycaffe

ineabstinencein

either

theplaceboor

caffe

ineconditions

31.Y

eomanset

al.(1998)e

N=36

caffe

ineconsum

ers(m

ean329mg/day);w

ithin

subjecta

ndacross

groupdesig

n;on

4days

subjectsabsta

ined

from

caffe

ineovernighta

ndweregivenbeverages(herbaltea

andfru

it-fla

voreddrinks,respectively)

containing

either

caffe

ine

(100

mg)

orplaceboat

breakfasta

ndmidmorning

(2hlater);

therewerefour

experim

entalg

roups:(1)b

reakfastcaffe

ine/

midmorning

caffe

ine,

(2)b

reakfastcaffe

ine/midmorning

placebo,

(3)b

reakfastplacebo/midmorning

caffe

ine,

(4)b

reakfast

placebo/midmorning

placebo;

abstinencenotb

iologically

verifiedf

Mooddata

werecollapsed

across

the4days;c

omparedwith

subjectswho

received

caffe

ine,

subjectswho

received

placebo

with

breakfasth

adlower

ratin

gsof

livelyandenergetic

2hlater;comparedwith

subjectswho

received

placeboat

breakfast

andcaffe

inemidmorning,subjectswho

received

placeboat

both

breakfasta

ndmidmorning

hadlower

ratin

gsof

energy

30min

afterthe

midmorning

beverage;the

authorsinterpretedthediffe

rences

betweencaffe

ineandplaceboas

refle

cting

caffe

ine-with

draw

alreversal

effects

32.D

ewset

al.(1999)

N=57

daily

coffe

edrinkers

(mean200–

300mg/daycaffe

ineim

mediately

before

abstinence);three

groups

ofsubjectsgiven

insta

ntcoffe

efor1

4days;the

chroniccaffe

inegroupreceived

insta

ntcaffe

inated

coffe

ethroughout

thestu

dy,the

abrupt

abstinencegroupreceived

insta

ntdecaffe

inated

coffe

efor7

consecutivedays,the

gradualabstin

ence

groupreceived

gradual

reductions

incaffe

ineover

5days;a

bstin

ence

was

biologically

verified

Nosta

tistical

analyses

werereporte

d;theauthorsconcludedthat

39%

ofsubjectsshow

edwith

draw

alin

thefirst2days

ofabstinencein

theabrupt

abstinencegroupbasedon

acompositewith

draw

almeasure

deriv

edfro

mamoodandattitude

questio

nnaire

andspontaneously

reporte

dwith

draw

alsymptom

s(e.g.,headache

andtiredness);22%

show

edsubstantial

decreases(≥1.5pointson

a4pointscale)intheirratings

ofdaily

functio

ning;o

nthefirst2days

ofabstinence28%

ofthe

abrupt

abstinencegroupreporte

dheadache,w

hile

17%

ofthechroniccaffe

inegroupreporte

dheadache

33.E

vans

andGriffiths

(1999)

[Exp

1]N=

15daily

caffe

ineconsum

ers(m

ean241mg/day)

with

in-subjectsdesig

n;subjectsweremaintainedon

capsules

containing

300mgcaffe

ine/day;

2-dayplaceboperio

dwas

repeated

sixtim

eswith

5–9days

ofcaffe

inebetweenperio

ds;abstin

ence

was

biologically

verified

Comparedwith

caffe

inedays,o

nplacebodays

subjectsreporte

dincreasedheadache,h

eadache/poor

mood,

tiredness,flu-like

symptom

s,fatig

ue(POMS),a

nddecreasedvigor(PO

MS),frie

ndlin

ess(POMS),a

ndactiv

ity/alertn

ess

34.E

vans

andGriffiths

(1999)

[Exp

2]N=

17daily

caffe

ineconsum

ers(m

ean277mg/day);w

ithin-subjectsdesig

n;2-dayplaceboperio

dafters

ubjectswere

maintainedon

capsules

containing

100,

300,

or600mgof

caffe

ine/day;

each

subjectw

asexposedto

two2-dayplacebo

perio

dsat

each

caffe

inedose;a

bstin

ence

was

biologically

verified

Comparedwith

caffe

inedays,o

nplacebodays

aftere

achmaintenance

dose

subjectsreporte

dincreasedheadache,fatigue

(POMS),confusio

n–bewild

erment(PO

MS),and

tiredness;the

rangeandmagnitude

ofwith

draw

alsymptom

sincreased

with

maintenance

dose,w

ithsubjectsreporting

thefollo

wingadditio

nalsym

ptom

safte

rthe

highestd

ose:increasedheadache/poor

mood,

totalm

ooddistu

rbance

(POMS),a

ndflu

-like

symptom

s,anddecreasedfriendliness(POMS),v

igor

(POMS),a

ndactiv

ity–alertn

ess;specificity

ofwith

draw

aleffectscanbe

concludedbecausehigher

dosesof

caffe

ineproduced

greater

response

toplacebowith

nochange

inbaselin

eresponse

tocaffe

ine

Tab

le1(con

tinued)

5

Reference

Design

With

draw

alsig

nsandsymptom

sb,c

35.E

vans

andGriffiths

(1999)

[Exp

3]N=

19daily

caffe

ineconsum

ers(m

ean294mg/day);w

ithin-subjectsdesig

n;subjectsweremaintainedon

capsules

containing

300mgcaffe

ine/day;

approxim

atelyevery7days

alower

dose

ofcaffe

ine(200,1

00,5

0,25

mg)

orplacebowas

substituted

for2

days;a

bstin

ence

was

biologically

verified

Comparedwith

thecaffe

inemaintenance

dose

of300mg,

placebosubstitutionproduced

increasedheadache,h

eadache/poor

mood,tiredness,fatigue

(POMS),and

totalm

ooddistu

rbance

(POMS)

anddecreasedvigor(PO

MS)

andactiv

ity/alertn

ess;the

rangeandmagnitude

ofwith

draw

alsymptom

sincreasedas

afunctio

nof

decreasin

gdosesof

caffe

ine,

with

nowith

draw

alreporte

dat

200mgandfatig

ue(POMS)

andtiredness

reporte

dat

100mg

36.E

vans

andGriffiths

(1999)

[Exp

4]N=

25daily

caffe

ineconsum

ers(263mg/day);w

ithin-subjectsd

esign;

subjectsmaintainedon

capsules

containing

300mg/day

fore

ither

1,3,

7,or

14consecutivedays;e

achduratio

nof

caffe

ineexposure

was

follo

wed

by7days

ofplacebo;

abstinence

was

biologically

verified

Nowith

draw

alsymptom

swereshow

nin

response

toplaceboafter1

dayof

caffe

ine;placeboafter3

days

ofcaffe

ineresultedin

increase

headache/poorm

ood,

fatig

ue(POMS),tire

dness,flu

-like

symptom

s,andtotalm

ooddistu

rbance

(POMS)

and

decrease

activ

ity/alertn

ess;placeboafter7

days

or14

days

ofcaffe

ineproduced

headache

inadditio

nto

mosto

fthe

symptom

sthatoccurre

din

the3-daycondition;specific

ityof

with

draw

aleffectscanbe

concludedbecauselonger

duratio

nsof

caffe

ine

exposure

produced

greaterresponseto

placebo

37.Jones

etal.(2000)

N=10

daily

caffe

ineconsum

ers(m

ean333mg/day);w

ithin-subjectsdesig

n;before

twosessions

subjectsabsta

ined

from

caffe

inefor2

1handthen

received

capsules

containing

caffe

ine(m

atched

totheiru

sualintake)o

rplacebo;subjectsw

erealso

evaluatedwhile

consum

ingtheirn

ormal

caffe

ineintake

(baseline);a

bstin

ence

was

biologically

verified

Comparedwith

caffe

ine,placebodecreasedalert/attentive/observant,ability

toconcentra

te,energy/activ

e,vigor(PO

MS),and

friendliness(POMS)

andincreasedheavyfeelings

inarmsandlegs;p

lacebo

also

increasedvelocitiesin

middleandanterio

rcerebral

arterie

sandincreasedEE

Gthetapower;c

affeineandbaselin

econditionswerenotsignific

antly

diffe

rent

onany

measure

38.S

werdlow

etal.(2000)

N=12

males

(mean173mg/day)

inthecaffe

ineabstinencestu

dyand18

males

(mean147mg/daycaffe

ine)

inthenon-

abstinencestu

dy;inthea

bstin

ence

study

subjectsabsta

ined

for>

14hpriortotesting;inboth

studies

subjectsreceived

either

apillcontaining

200mgcaffe

ineor

placebo;

abstinencenotb

iologically

verified

Comparedwith

caffe

ineadministratio

n,placeboadministratio

nto

high

caffe

ineconsum

ers(asdeterm

ined

bymediansplit)

produced

signific

ante

levatio

nsin

acompositesomatic

symptom

scale(fe

elsic

k,queasy,d

izzy,a

ndperspirin

g).S

imilar

effectswereshow

nforthe

individual

symptom

s39.Y

eomanset

al.(2002b)

N=30

caffe

ineconsum


ithin

subjectsandacross

groupdesig

n;aftero

vernight

abstinencethreegroups

ofsubjectsreceived

placebo,1-mg/kg,or2

-mg/kg

caffe

inein

aflavoreddrinkatbreakfasto

n2testdays;6

0min

afterthe

first

dose

subjectsreceived

adrinkcontaining

either

placeboor

1-mg/kg

caffe

inein

counterbalancedordera

crossthe

2testdays;

abstinencenotb

iologically

verified

Comparedwith

subjectsgivencaffe

ine,

subjectsgivenplaceboat

breakfasth

adslo

wer

reactio

ntim

esandlower

alertness

45min

later;relativ

eto

placebo,

caffe

inegiven60

min

afterb

reakfastdecreasedreactio

ntim

eandincreasedalertnessin

subjectswho

received

placeboatbreakfast,butn

otin

subjectswho

received

caffe

inea

tbreakfast;

thea

uthorsinterpretedthese

results

assuggestin

gthat

reversal

ofcaffe

inewith

draw

alisamajor

component

oftheeffectsof

caffe

ineon

moodand

perfo

rmance

40.R

ogerset

al.(2003)e[Exp

1]N=

10caffe

ineconsum

ers(m

ean355mg/day)

and10

non-consum

ers;betweengroups

desig

n;before

four

sessions

subjects

absta

ined

from

caffe

inefora

pproximately15

h;abstinencenotb

iologically

verifiedf

Afte

rabstin

ence,c

affeineconsum

erswereless

alerta

ndmoretensethan

non-consum

ers

41.R

ogerset

al.(2003)e[Exp

2]N=

22caffe

ineconsum

ers(m

ean372mg/day)

and20

non-consum

ers;betweengroups

desig

n;aftero

vernight

abstinence

subjectsreceived

either

100mgcaffe

ineor

placeboin

afla

voreddrink;

abstinencenotb

iologically

verifiedf

Afte

rabstin

ence,c

affeineconsum

erswereless

alertthannon-consum

ers

42.T

inleyet

al.(2003)e

N=45

caffe

ineconsum

ers(m

ean324mg/day);w

ithin

subjectsandacross

groups

desig

n;for2

-weeks

subjectsabsta

ined

from

theirn

ormalsourceso

fcaffeineandconsum

edteaandcoffe

eprovided

bytheexperim

enters;inonegrouptheteaandcoffe

ewerecaffe

inated

andin

theotherg

roup

they

werecaffe

inefre

e;subjectsabsta

ined

from

teaandcoffe

eovernightb

eforeeach

offour

testsessions

conductedover

the2-weekperio

d;subjectsreporte

dmoodbefore

andafterreceiving

asin

gleservingof

either

acaffe

inebeverage

(caffeinemaintainedgroup)

orcaffe

ine-fre

ebeverage

(caffeine-fre

egroup);a

bstin

ence

was

biologically

verified

Comparedwith

thegroupthat

was

chronically

maintainedon

caffe

ine-fre

ebeverages,thegroupthat

was

maintainedon

caffe

inereporte

dhigher

levelsof

headache

aftero

vernight

abstinence;

headache

was

alleviated

afterc

affeineadministratio

n

a Allstud

ieshadacaffeine-abstin

ence

period

of12

hor

greater

bAlleffectsarestatistically

sign

ificantat

P≤0

.05un

less

otherw

iseno

ted

c For

consistency,

whenplaceboandcaffeine

cond

ition

sarecompared,

theeffectsaredescribedas

theeffectsof

placeborelativ

eto

caffeine

dPOMSindicatesProfile

ofmoo

dstates

questio

nnaire

e Onlythosedesign

elem

entsthat

arerelevant

totheanalysisof

caffeine

abstinence

effectsaredescribed

f Alth

ough

abstinence

was

notbiolog

ically

verified,saliv

asamples

weretakenandsubjectswereledto

believe

that

samples

wou

ldbe

analyzed

forcompliance

gLikelyadirect

effect

ofcaffeine

rather

than

aneffect

ofcaffeine

with

draw

alhSom

eof

theresults

arebasedon

authors’

conclusion

sbecauseno

tallsign

ificantcomparisons

werepresented(Lader,person

alcommun

ication,

Aug

ust20

02)

Tab

le1(con

tinued)

6

Tab

le2

Sum

maryof

sing

le-blin

dandno

n-blindexperimentalstud

iesof

caffeine

with

draw

ala

Reference

Design

With

draw

alsig

nsandsymptom

sb,c

Single-blin

d43.D

riesbachandPfeiffe

r(1943)

N=22

adults;

with

in-subjectsd

esign;

caffe

ineadministered

incapsules

inincreasin

gdoseso

ver6

–7days

to650–

780mg/day;

onthe7th

or8thdayplacebowas

substituted

forc

affeinecapsules;a

bstin

ence

notb

iologically

verified

Placebosubstitutionafterc

affeineresultedin

lethargy

inmorning,cerebralfullnessatnoon,and

headache

inearly

afternoon,

reaching

peak

intensity

3–6hlater;82%

ofsubjectsreporte

ddefin

iteto

severe

headache;in33%

ofthesesubjectsheadache

was

accompanied

bynausea

andsometim

esvomiting

andserio

usrhinorrhea;in55%

oftrialsheadache

was

reporte

dto

beas

severe

asthesubjecth

adeverexperie

nced;h

eadachew

asalleviated

byre-adm

inistratio

nof

caffe

ine;otherw

ithdraw

alsymptom

swerenotedbutn

otsta

tistically

analyzed

(i.e.,m

entald

epression,

drow

siness,yawning,a

nddisin

clinationto

work);s

erum

protein,

serum

calcium,a

ndhematocrit

weresig

nific

antly

lower

durin

gheadache

andserum

inorganicphosphorus

was

slightly

higher

44.H

öfer

andBä

ttig

(1994a)

N=120habitualcoffe

eusers(mean5.7cups/day);betweengroups

desig

n;aftera

3-daybaselin

eperio

d,subjectsreceived

either

9days

ofcaffe

inated

insta

ntcoffe

e,9days

ofdecaffe

inated

insta

ntcoffe

eor

9days

ofinterm

ittentc

affeinated

anddecaffe

inated

insta

ntcoffe

e;abstinencewas

biologically

verified

Intheinterm

ittentg

roup,c

omparedwith

caffe

inedays,d

ecaffeinated

days

produced

increasednausea

andindisposition;inthe

decaffe

inated

group,

decaffe

inated

days

comparedwith

baselin

eproduced

increasedheadache,u

seof

analgesic

s,indisposition,h

eart

rate,and

decreasedwakefulness,w

ell-b

eing,d

aypositive,motor

activ

ity,and

problemsfallin

gasleep;inthedecaffe

inated

groupthese

effects(exceptfor

heartrate)

peaked

ondays

1or

2andprogressivelydecreasedtowardpre-with

draw

allevels;

subjectsin

the

interm

ittentg

roup

show

edmosto

fthese

effects;40

–50%

ofsubjectsreporte

dmoderateheadache

indecaffe

inated

groupand20–3

0%reporte

dmoresevere

headache

45.H

öfer

andBä

ttig

(1994b)

N=42

femalehabitual

coffe

edrinkers

(mean6cups/day);betweengroups

desig

n;aftera

3-dayphaseof

usualc

offeeconsum

ption

(baseline),subjectsenteredasecond

phaseconsistingof

either

3days

ofdecaffe

inated

insta

ntcoffe

eor

3days

ofcaffe

inetablets;

abstinencewas

biologically

verified

Caffe

inewith

draw

alwas

defin

edas

asta

tistical

interactionbetweengroupandphase,

with

thedecaffe

inated

coffe

egroupshow

ing

increasedheadache,m

uscle/jointa

che,

sleep

duratio

nanddecreasedwell-b

eing,ratings

ofdaypositive,

wakefulness,a

ndmotor

activ

ity;stomach/belly

ache

was

increasedin

thecaffe

inetablet

groupanddecreasedin

thedecaffe

inated

coffe

egroupd;d

esire

for

coffe

ewas

increasedin

thecaffe

inetablet

groupbutn

otchangedin

thedecaffe

inated

coffe

egroup

46.L

ane(1994)

N=14

habitual

coffe

edrinkers

(3.4

cups/day);with

in-subjectsdesig

n;aftero

vernight

abstinencesubjectsweregivenwater

containing

300mgcaffe

ineor

placebo;

abstinencenotb

iologically

verified

Comparedwith

caffe

ine,placeboproduced

increaseddrow

sy/sleepy,lethargy/fatig

ue/tired/sle

epy,headache,self-c

onfid

ence,and

heavy

feelings

inarmsa

ndlegs,and

decreaseddesireto

socialize/talkativeness;9of

14subjectsreporte

dmoreheadache

afterp

lacebo

andno

subjectsreporte

dmoreheadache

afterc

affeine;

placebowas

associated

with

decreasedlevelsof

urinaryepinephrinewhich

were

interpretedby

theauthor

asan

effect

ofcaffe

inerather

than

ofcaffe

inewith

draw

al47.R

ubin

andSm

ith(1999)

N=43

regularc

offeeor

teaconsum

ers(m

ean175mg/day);w

ithin-subjectsdesig

n;aftera

2-daybaselin

eperio

dof

norm

alcaffe

ine

consum

ption,

subjectsconsum

edcaffe

inated

coffe

e(ortea)for

2days

anddecaffe

inated

coffe

e(ortea)for

2days

incounterbalanced

order;abstinencenotb

iologically

verified

Asig

nific

antly

greatern

umbero

fsubjectsreportedheadache

durin

gthedecaffe

inated

perio

d(40%

)thandurin

gthebaselin

e(14%

)and

caffe

inated

perio

d(19%

);thesediffe

rences

weregreaterinindividualswho

corre

ctly

identifiedthepresence

orabsenceof

caffe

ine

(49%

indecaffe

inated

perio

d,7%

inbaselin

e,9%

incaffe

inated

condition)

48.F

ield

etal.(2003)

N=10

lowcaffe

ineconsum

ers(mean41

mg/day)

and10

high

caffe

inec

onsumers(mean648mg/day);w

ithin

subjectsandacross

groups

desig

n;before

twosessions,subjectsabsta

ined

from

caffe

inefor3

0h;

cerebral

bloodflo

wwas

assessed

with

perfu

sionmagnetic

resonanceim

aging60–90min

afterreceiving

capsules

containing

caffe

ine(250

mg)

orplacebo;

abstinencenotb

iologically

verified

Comparedwith

caffe

ine,

placeboproduced

greaterc

erebralb

lood

flow

velocitiesin

white

matter,anterio

rgraymatter,andposte

rior

gray

matterregions;h

ighcaffe

ineuserss

howed

agreaterc

affeine–placebodiffe

rencein

theanterio

rgraymatterthanlowusers;across

allsubjects,cerebral

bloodflo

win

theplacebocondition

increasedlin

early

with

daily

caffe

ineintake

Non-blin

d49.N

aism

ithet

al.(1970)

N=20

caffe

ineconsum

ers(560

mg/day);w

ithin-subjectsdesig

n;aftera

10-day

baselin

eperio

d,subjectssw

itchedto

decaffe

inated

coffe

eandabsta

ined

from

allo

ther

sourcesof

caffe

inefor1

4days;a

bstin

ence

notb

iologically

verified

Nosta

tistical

analyses

werereporte

d;allsubjectsreporte

dlassitu

deandsevere

headache

which

resolved

by48

h

50.R

oller(1981)

N=1malehabitualcoffe

euser

(900–1,100

mg/day);w

ithin-subjectsd

esign;

absta

ined

from

caffe

inefor7

2-hperio

d;24

hinto

this72-h

perio

dtheophyllin

ewas

administered;a

tthe

endof

72heither

caffe

inated

coffe

e(approx.

150mgcaffe

ine)

ordecaffe

inated

coffe

ewas

ingeste

din

ablindedfashion

Nosta

tisticalanalyses

werereporte

d;headache

starte

daftera

bout

6h;

shortly

thereafte

rthe

subjectreportedtiredness

orlassitu

de,then

rhinorrhea

andlegmusclepains,follo

wed

bydiaphoresis

;afte

r16hthesubjectreportedgeneralm

usclepains;thesesymptom

sgradually

increasedto

amaxim

umintensity

atalatertim

eandweresuppressed

bycaffe

inated

coffe

econsum

ptionbutn

otplacebo

51.E

delsteinet

al.(1983)

N=18

psychiatric

patientsw

hoconsum

edthreeor

morecaffe

inated

beveragesp

erday;

with

in-subjectsd

esign;

decaffe

inated

beverages

weresubstituted

forc

affeinated

beverages;abstinencenotb

iologically

verified

Nosta

tistical

analyses

werereporte

d;headache

increaseddurin

gthefirst1–2weeks

afterd

ecaffeinated

beveragesweresubstituted

for

caffe

inated

beverages

52.M

athew

andWilson

(1985)

N=8heavy(m

ean986mg/day)

and8lig

ht(m

ean126mg/day)

caffe

ineusers;with

in-subjectsd

esign;

afterb

aselineassessmentsubjects

absta

ined

from

caffe

inefor2

4h;

abstinencenotb

iologically

verified

Comparedwith

cerebral

bloodflo

wmeasurementsassessed

after2

-habstinence,

24-h

abstinenceproduced

increasedcerebral

blood

flow

inseveralfrontal

regionsbilaterally

inheavyusersbutn

otlig

htusers

53.R

izzo

etal.(1988)

N=20

daily

caffe

ineusers(m

ean606mg/day)

andN=

20caffe

inenon-users;with

in-subjectsdesig

n;twoexperim

entalsessio

nsseparatedby

7days;u

sers

absta

ined

from

caffe

inefor2

days

priortothesecond

session;

abstinencenotb

iologically

verified

Caffe

ineusershadslo

wer

reactio

ntim

esandsm

allerh

eartratedeceleratio

nsthan

non-usersdurin

gcaffe

ineabstinence(sessio

n2)

but

nota

tbaseline(sessio

n1)

54.C

outurie

retal.(1997)

N=20

adultspresum

edto

bedaily

caffe

ineusers;with

in-subjectsdesig


atbaselin

e,24

hlatera

fterc

omplete

caffe

ineabstinence,and30

min

and120min

afterreceiving

capsules

containing

150mgcaffe

ine;abstinencewas

biologically

verified

Subjectsreporte

dheadache

aftera

bstin

ence

with

completerecovery

afterc


n;comparedwith

baselin

ecerebral

bloodflo

wvelocities(le

ftmiddlecerebral,b

asilar,andboth

poste

riorc

erebrala

rterie

s)increasedaftera

bstin

ence;c

omparedwith

abstinencecerebral

bloodflo

wvelocitiesdecreasedafterc


n;50%

ofsubjectsreporte

dmoderateto

severe

headache;1

0%reporte

dnausea

andvomiting

55.L

ane(1997)

N=16

daily

caffe

ineconsum

ers(m

ean612mg/day);w

ithin-subjectsdesig


ontwooccasio

ns:a

fteru

sual

caffe

ineconsum

ptionandaftero

vernight

abstinence(estimated

12–2

8habstinence);a

bstin

ence

notb

iologically

verified

Comparedwith

adlib

itum

caffe

ineconsum

ption,

caffe

ineabstinenceincreasedheadache,d

rowsy/sleepy,fatigue

(POMS)

e ,lethargy/

fatig

ue/tired/slu

ggish

,muzzy/fo

ggy/notclearheaded,y

awning,flu-like

feelings,lightheaded/dizzy,heavy

feelingin

armsa

ndlegs,and

decreasedvigor(PO

MS),w

ell-b

eing,d

esire

tosocialize/talkativeness,ability

toconcentra

te,e

nergy/activ

e,content/satisfied,

mean

arteria

lblood

pressure;a

ftera

bstin

ence

63%

ofsubjectsreporte

dheadache,2

5%reporte

dmaxim

umratin

gsof

headache,a

nd31%

reporte

dflu

-like

symptom

s

7

Reference

Design

With

draw

alsig

nsandsymptom

sb,c

56.L

aneandPh

illips-

Bute

(1998)

N=30

daily

coffe

econsum

ers(mean569mg/daycaffe

ine);w

ithin-subjectsd

esign;

subjectswereassessed

ontwooccasio

ns:afte

rusual

caffe

ineconsum

ptionandaftero

vernight

abstinence(estimated

12–2

8habstinence);a

bstin

ence

notb

iologically

verified

Comparedwith

adlib

itum

caffe

ineconsum

ption,caffe

ineabstinenceincreasedheadache,drowsy/sleepy,irritable/cross/g

rumpy;fatigue

(POMS),lethargy/fatig

ue/tired/slu

ggish

,muzzy/fo

ggy/notc

learheaded,y

awning,a

nger–hostility

(POMS),c

onfusio

n–bewild

erment

(POMS),d

epression-dejection(POMS),a

nddecreasedvigor(PO

MS),w

ell-b

eing,d

esire

tosocialize/talkativeness,ability

toconcentra

te,energy/activ

e,alert/attentive/observant;urge

todo

task/work-relatedactiv

ities;h

otor

cold

spells,

need

topass

water

d ;on

avisualvigilancetask

abstinencedecreasedhit-rateandincreasedresponse

time,andabstinenceincreasedsubjectsratin

gsof

perceived

difficulty

anddecreasedperceivedim

porta

nceof

doingwella

ndperceivedsuccess;aftera

bstin

ence

47%

reporte

danyheadache

and

27%

reporte

dmoderateor

severe

headache

57.R

eeveset

al.(1995,

1997,2002);P

atric

ketal.

(1996)

N=13

or14

daily

caffe

ineusers(>300mg/day);w

ithin-subjectsdesig

n;subjectsevaluatedwhile

consum

ingtheirn

ormal

caffe

inediet

(baseline)

andagainafter1

,2,a

nd4days

ofabstinence;

evaluatio

nsincluded

topographicquantitativeEE

G,p

hysic

ianratin

gsof

caffe

inewith

draw

alseverityas

assessed

durin

gastructuredinterview,

andamusculoskeletalexam

inationof

thespine;abstinencewas

biologically

verified

Comparedwith

baselin

e,sig

nific

antq

uantitativ

eEE

Gchangesw

ere:increase

inthetaabsolutepower,increasein

delta

absolutepower

over

thefro

ntalcorticalareas,d

ecreasein

themeanfre

quency

ofboth

alphaandbetarhythm

,increasein

thetarelativ

epower,decrease

inbeta

relativ

epower,and

change

ininterhem

ispheric

coherence;

resumptionof

caffe

inefollo

wingabstinencereturned

alteredEE

Gvalues

tobaselin

elevels;

allsubjectsreportedcaffe

ine-with

draw

alsymptom

sand

77%

hadmoderateor

severe

with

draw

alseverityas

ratedby

aphysician;

comparedwith

baselin

e,thenumbero

fsom

atic

dysfunctions

from

themusculoskeletal

exam

inationwas

signific

antly

greatero

ndays

1,2,and4of

caffe

ineabstinence;thegreatestnumbero

fsom

aticdysfunctions

was

reporte

don

day2;

sixsubjectswho

hadbaselin

eEE

Gscontaining

diffu

separoxysm

alslo

wing(D

PS,a

minor

EEG

dysrhythmia)s

howed

signific

ant

increasesin

DPS

firingdurin

gwith

draw

alandareturn

tobaselin

efiringlevelsfollo

wingcaffe

ineresumption

a Allstud

ieshadacaffeine

abstinence

period

of12

hor

greater

bAlleffectsarestatistically

sign

ificantat

P≤0

.05un

less

otherw

iseno

ted

c For

consistency,

whenplaceboandcaffeine

cond

ition

sarecompared,

theeffectsaredescribedas

theeffectsof

placeborelativ

eto

caffeine

dLikelyadirect

effect

ofcaffeine

rather

than

aneffect

ofcaffeine

with

draw

ale POMSindicatesProfile

ofmoo

dstates

questio

nnaire

Tab

le2(con

tinued)

Tab

le3

Sum

maryof

survey

stud

iesof

caffeine

with

draw

al

Reference

Design

With

draw

alsig

nsandsymptom

s

58.G

oldstein

and

Kaizer(1969)

N=183femalecoffe

econsum

erssurveyed

aboutthe

effectsthey

would

experie

nceifmorning

coffe

ewas

omitted

Endorsem

ento

fsym

ptom

sincreased

asthenumbero

fcupso

fcoffeeconsum

edincreased;

amongheavycoffe

eusers(5–

10cups/day)the

percentagesreporting

specificsymptom

swere58%

feel

halfaw

ake,

24%

lethargic,

21%

irritable,

18%

sleepy,

16%

unable

towork

effectively,

15%

restless,8%

headache

59.W

inste

ad(1976)

N=135mostly

adultinpatientson

psychiatric

ward;

survey

ofoccurre

nceof

caffe

ine-with

draw

alsymptom

s26%

ofheavycaffe

ineusers(>500mg/day)

reporte

dexperie

ncing“anxiety

with

draw

alsymptom

s”60.G

redenet

al.

(1978)

N=83

psychiatric

inpatients;surveyed

aboutthe

occurre

nceof

headache

onom

issionof

morning

coffe

e11%

endorsed

experie

ncingheadache

61.G

redenet

al.

(1980)

N=205medical

inpatients;questio

nedaboutthe

occurre

nceof

headache

upon

stoppingcaffe

ine

20%

oftotalsam

ple(or2

8%of

thesampleof

152afterthose

who

answ

ered

“don

’tknow

”wereexcluded)reportedcaffe

ine-with

draw

alheadache;those

reporting

headache

hadhigher

meancaffe

ineintake

(616

mg/day)

than

thosenotreportin

gheadache

(395

mg/day)

62.V

ictore

tal.(1981)

N=124generalm

edical

inpatients;survey

oftheoccurre

nceof

caffe

ine-with

draw

alheadache

24%

ofsubjectsendorsed

having

experie

nced

caffe

ine-with

draw

alheadache

63.H

ugheset

al.

(1998)

N=162caffe

ineconsum

ers;surveyed

viatelephoneaboutc

affeineabstinenceandwith

draw

alsymptom

s44%

ofcaffe

ineconsum

ersreporte

dhaving

absta

ined

from

caffe

inefor≥

24hin

thepastyear;o

fthose,2

4%reporte

dheadache,2

7%drow

siness,21%

fatig

ue,1

0%anxiety,4%

depressio

n,3%

nausea

orvomiting,2

8%strongdesireto

use,20%

irritability,21%

yawning,

11%

difficulty

concentra

ting,

18%

less

motivated

towork

64.D

ewset

al.(1999)

N=6,815adultswho

reporte

ddaily

caffe

ineusewereaskedifthey

had“problem

sor

symptom

son

stoppingcaffe

inein

thepast”

11%

reporte

dwith

draw

alsymptom

supon

stoppingcaffe

ine;

ofthose,

25%

reporte

dsymptom

ssevere

enough

tointerfe

rewith

norm

aldaily

activ

ities

(e.g.,losttim

efro

mwork);h

eadachewas

themostc

ommon

symptom

reporte

d;itisunknow

nwhatp

ercentageof

daily

usershadactually

absta

ined

from

caffe

ine

65.K

endler

andPre-

scott(1999)

N=1,642wom

enin

apopulatio

n-basedtwin

registryinterviewed

abouttheircaffe

ineuse;

individualswho

reporte

dsto

ppingor

cutting

downtheirc

affeineconsum

ptionwereaskedaboutc

affeine-with

draw

alsymptom

s24%

ofsubjectswho

reporte

dsto

ppingor

cutting

downtheirc

affeineconsum

ptionreporte

dhaving

experie

nced

headache

plus

oneor

moreof

thefollo

wing:

markedfatig

ueor

drow

siness,markedanxietyor

depressio

n,andnausea

orvomiting

66.B

ernstein

etal.

(2002);O

berstare

tal.(2002)

N=36

adolescent

daily

caffe

ineconsum

ers(m

ean244mg/day)

who

endorsed

twoor

moreDSM

-IVcaffe

inedependence

criteria

durin

gtelephoneprescreening;interview

edandassessed

forc

affeinedependence

basedon

DSM

-IVsubstancedependence

criteria

;21of

these

subjectswerereassessed

1year

later

Basedon

both

interviews,81%

metthew

ithdraw

alcriterio

n;100%

ofsubjectswho

metcriteria

forD

SM-IV

caffe

inedependence

reporte

dwith

draw

alsymptom

s;thepercentage

ofsubjectsreporting

specificsymptom

swere59%

drow

sy/tired,

56%

fatig

ued,

56%

sluggish

/slo

wed

down,

56%

headache,3

2%restless/c

annotsitstill,

29%

nervous/a

nxious,2

1%sic

k/nauseated/vomiting,2

1%drinkcaffe

ineso

youdo

notfeelb

ad,a

nd9%

sad/depressed

8

Table 4 Summary of withdrawal symptoms reported in experimental and survey studies described in Tables 1, 2, 3a

Symptom Acute abstinenceversus baselineb

Acute abstinenceversus caffeineb

Acuteabstinencein caffeineconsumersversus non-consumersb

Acute abstinence versuschronic abstinenceb

Time-limitedabstinence effectsb

Percentage of subjects showing effect inexperimental studiesc

Percentage of subjects reportingsymptom in survey studiesd

Headache 4, 9, 20, 44, 45, 47,49, 50, 51, 54

1, 2, 5, 6, 7, 8, 9, 10,11, 12,13, 15, 16, 21, 23,25, 27, 30,33, 34, 35, 36, 43,46, 47, 55, 56

17 6, 8, 27, 42 4, 5, 7, 44, 49, 51 1: 25% of trials (heavy users) [control 3%] 58: 8% headache4: 100% headache (control 14%) 60: 11% headache6: 79% headachee* 61: 20% headache7: 42% headache (control 10%) 62: 24% headache9: 52% moderate or severe headache (control2%)

63: 24% headache

11: 30% headache* 65: 24% headache plus additionalsymptom

12: 64% maximum headache (control 0%) 66: 56% headache15: 9% headache*

20: 45% diffuse throbbing headache

27: 47% headache (control 14%)


43: 82% definite to severe headache, 55% oftrials severe as ever experienced

44: 40–50% moderate headache, 20–30% moresevere headache



49: 100% headache

54: 50% moderate or severe headache

55: 63% headache (control 6%), 25% maximumheadache (control 0%)

56: 47% headache, 27% moderate or severeheadache

Tiredness/fatigue 4, 9, 20, 49, 50 2, 5, 6, 8, 9, 11, 12,14, 15,16, 17, 18, 19, 21,23, 24,25, 26, 28, 30,

33, 34, 35, 36, 43,46, 55, 56

17 6, 8, 17, 26 4, 5, 17, 20, 49 6: 57% lethargy/fatiguee* 58: 24% lethargic9: 8% abnormal high fatigue (control 0%) 63: 21% fatigue11: 24% fatigue* 66: 56% fatigued12: 27% extreme fatigue (control 9%)

15: 18% fatigued/tired*

20: 68% tired and lethargy

49: 100% lassitude

Decreased energy/ac-tiveness

4, 9 2, 5, 6, 8, 9, 12, 15,18, 19,21, 22, 23, 24, 26,

28, 29, 31, 33, 34,35, 37, 55, 56

2 6, 26 4, 5 6: 64% ↓ energy/activee*

9: 11% abnormal scores (control 0%)

12: 45% extreme ↓ in vigor (control 0%)

15: 27% ↓ stimulated/active/energy/excited*

Decreased alertness/attentiveness

4, 20 2, 5, 6, 8, 15, 19, 21,24, 25,26, 27, 33,

34, 35, 36, 37, 39,56

2, 40, 41 6, 26 4, 5 6: 50% ↓ alert/attentivee*

15: 27% ↓ alert/attentive/able to concentrate*

Drowsiness/sleepi-ness

4, 9, 20, 44, 45 2, 6, 8, 9, 11, 14, 15,17, 19,21, 23, 24, 28, 46,55, 56

2, 17 6, 17, 26 4, 44 6: 64% drowsy–sleepye* 58: 18% sleepy, 58% feel-half awake11: 24% drowsiness* 63: 27% drowsiness15: 45% drowsy/sleepy/tired* 66: 59% drowsy/tired22: 63% drowsiness/fatigue*

23: 13% drowsiness*

Decreased contented-ness/well-being

9, 44, 45 2, 5, 6, 9, 17, 18, 19,21, 22,23, 24, 25, 26, 55,56

2 6, 26 5, 44 6: 64% ↓ well-beinge*, 50% ↓ content/satisfiede*

15: 18% ↓ content/relaxed/satisfied*

Decreased desire tosocialize

4, 9 2, 5, 6, 9, 10, 14, 17,22, 33,34, 37, 46, 55, 56

6 5 6: 79% ↓ social dispositione*

15: 9% ↓ talkative*

“Flu-like” symptoms 9 5, 6, 9, 25, 30, 33,34, 36, 55

6 5 55: 31% flu-like symptoms (control 0%)

Depressed mood 9 5, 6, 9, 12, 14, 15,17, 56

17 6, 17, 26 5 6: 14% depressede 63: 4% depression9: 11% abnormal scores BDI (control 2%) 66: 9% sad/depressed12: 36% elevations on BDI (control 0%)

15: 18% depressed*

9

data on the independence of specific caffeine-withdrawalsymptoms (e.g., factor analysis), the listing provides auseful framework for characterizing the results of caffeine-withdrawal studies. In Table 4, symptoms are sequencedfrom those documented (i.e., shown in columns 2–6) in thegreatest number of experimental studies to those docu-mented in the fewest. Table 4 also differentiates experi-ments according to five of the most frequently used

experimental methodologies that have been used to drawinferences about caffeine-withdrawal symptoms and signs.Finally, Table 4 shows the percentages of individualsreporting withdrawal symptoms in both experimental andsurvey studies.

Symptom Acute abstinenceversus baselineb

Acute abstinenceversus caffeineb

Acuteabstinencein caffeineconsumersversus non-consumersb

Acute abstinence versuschronic abstinenceb

Time-limitedabstinence effectsb

Percentage of subjects showing effect inexperimental studiesc

Percentage of subjects reportingsymptom in survey studiesd

Difficulty concentrat-ing

20 5, 6, 19, 26, 37, 55,56

6, 26 5 6: 79% ↓ able to concentratee* 63: 11% difficulty concentrating

Irritability 9 2, 5, 6, 9, 19, 22, 56 2 6, 26 5 6: 29% irritable/cross/grumpye* 58: 21% irritable15: 9% irritable/frustrated/angry/cross* 63: 20% irritability

Unmotivated forwork

9 5, 6, 9, 21, 22, 23,24, 56

5 6: 57% ↓ motivation for worke* 58: 16% unable to work effectively

6: 64% ↓ urge to do worke* 63: 18% ↓ motivated to work

Muzzy/foggy/notclearheaded

9, 20 6, 9, 17, 18, 55, 56 17 6 6: 71% muzzye*

Yawning 9 6, 9, 19, 21, 28, 55,56

6 6: 43% yawninge* 63: 21% yawning

Decreased self-confi-dence

9 5, 6, 9, 21, 23, 24, 46 6 5 6: 64% ↓ self-confidencee*

Confusion–bewilder-ment

9 6, 9, 22, 26, 34, 56 26

Total mood distur-bance

9 6, 9, 26, 34, 35, 36 26

Nausea/vomiting/upset stomach

22, 23, 24, 38, 43, 44 6: 29% upset stomache* 63: 3% nausea or vomiting

20: 13% nausea and sickness with headache 66: 21% sick/nauseated/vomiting

43: 33% nausea (sometimes vomiting) alongwith headache

54: 10% nausea and vomiting

Muscle pain/stiffness 45, 50 5, 6 6 5 6: 43% muscle paine*

Anxiety/nervousness 9 9, 22 2, 40 9: 8% abnormal scores anxiety (control 0%) 59: 26% anxiety withdrawalsymptoms

63: 10% anxiety

66: 29% nervous/anxious

Heavy feelings inarms and legs

9 9, 37, 46, 55

Increased sleep dura-tion/quality

20, 44, 45 27

Analgesics use 44 9, 12 44 9: 13% analgesics use (control 2%)

12: 45% analgesics use

Craving 5, 6 6 5 6: 43% craving for caffeine* 63: 28% strong desire to use

Lightheaded/dizzy 6, 38, 55

Blurred vision 9 6, 9 6 6: 14% blurred visione*

Anger/hostility 17, 56 17 17

Hot and cold spells 9 9, 56

aNumbers in table refer to the entry number in Tables 1, 2, 3bThe strengths and weaknesses of each of the types of comparisons presented in columns 2–6 are described in the text. Studies in columns4–6 are more rigorous because the effects are not confounded by the direct effects of caffeinecThis column (column 7) shows the percentage of subjects reporting the symptom in experimental studies (as summarized in columns 2–6).For comparison, when available, the percentage of subjects reporting the symptom in the control condition (i.e., baseline or caffeinecondition) is shown in brackets. For studies that conducted within-subject analyses of differences between the abstinence and controlconditions, an asterisk (*) indicates the percentage of subjects who showed a statistically significant difference at P<0.05dPercentage of subjects reporting symptom in survey studies: this column shows the percentage of individuals reporting the occurrence of asymptom presumably based on their previous experience in the natural environmenteIncidence data for this study were collapsed across two conditions: acute abstinence versus caffeine and acute abstinence versus chronicabstinence

Table 4 (continued)

10

Experimental methodologies

Several different experimental methodologies have beenused to draw inferences about the occurrence of caffeinewithdrawal. As understanding the strengths and weak-nesses of each of these methodologies is critical todrawing meaningful conclusions about the validity of thefindings, this section will briefly list and critique the mostcommonly used experimental methodologies. The firstfive experimental methodologies (corresponding to col-umns 2–6 in Table 4) involve different comparisons withacute caffeine abstinence.

Acute abstinence versus preceding caffeine baseline(not counterbalanced)

In this within-subjects comparison, symptoms or signsduring acute caffeine abstinence are compared with thoseduring a preceding baseline of ad libitum caffeineconsumption. This comparison may have good ecologicalvalidity because it can involve a naturally occurringpattern of caffeine consumption followed by abruptabstinence. A limitation of this comparison is that baselineversus abstinence differences may be due to the simpleabsence of the direct effects of caffeine during abstinence(i.e., a caffeine offset effect rather than a time-limitedwithdrawal effect). Another limitation is that observeddifferences could be confounded by order effects (i.e.,conditions are not counterbalanced).

Acute abstinence versus caffeine

In this comparison, which could involve within subjects oracross groups designs, symptoms or signs during acuteabstinence are compared with those during caffeineconsumption, with conditions counterbalanced or rando-mized across subjects. As with the comparison with apreceding baseline caffeine condition, this comparisonmay have good ecological validity in modeling naturallyoccurring effects of caffeine abstinence. Although notconfounded by order effects, this comparison has thelimitation that caffeine versus abstinence differences maybe due to the simple absence of the direct effects ofcaffeine during abstinence.

Acute abstinence in caffeine consumers versus non-consumers

In this between-groups comparison, symptoms or signsduring acute abstinence in caffeine consumers arecompared with those in non-consumers. As both groupsare caffeine abstinent, it can be concluded that anydifference observed is not confounded by the direct effectsof caffeine. However, the possible confounding effects ofpopulation differences between caffeine consumers and

non-consumers cannot be ruled out because of the self-selected nature of these groups.

Acute abstinence versus chronic abstinence

In this comparison, which could be within subjects oracross groups, symptoms or signs during acute abstinenceare compared with those during chronic abstinence (e.g.,1 week or more of caffeine abstinence). As both conditionsinvolve caffeine abstinence, it can be concluded that anydifference observed is not confounded by the direct effectsof caffeine. Although this comparison provides moreconclusive evidence of withdrawal effects than thepreceding comparisons, the approach is conservativebecause it may underestimate the incidence or magnitudeof withdrawal effects if symptoms or signs persist in thechronic abstinence condition.

Time-limited abstinence effects

By definition, a drug withdrawal symptom or sign shouldincrease upon acute abstinence and then decrease overtime with continued drug abstinence. A demonstration ofsuch time-limited effects is not confounded by the directeffects of caffeine and is important for confirming that theeffects observed are withdrawal related rather thanreflecting caffeine offset effects (i.e., a return to thenormal drug-free state).

Other methodologies

In addition to the five approaches described above, severalother types of experimental methodologies can help toinform the interpretation of caffeine-withdrawal effects.Although not included in Table 4, studies using these othermethodologies are discussed in the following section onsymptoms and signs of caffeine withdrawal.

Variation in caffeine maintenance dose Demonstration thatthe severity or incidence of a symptom or sign increaseswith increases in the daily caffeine maintenance dosebefore abstinence helps to confirm that the withdrawaleffect reflects a pharmacological process.

Acute decreases in caffeine maintenance dose When lowercaffeine doses are substituted for the usual caffeinemaintenance dose, the demonstration that the severity orincidence of a symptom or sign increases as the substituteddose decreases helps to confirm that the withdrawal effectreflects a pharmacological process.

Manipulation of duration of caffeine maintenance De-monstration that the severity or incidence of a symptom orsign increases with increases in the duration of dailycaffeine maintenance dose before abstinence helps toconfirm that the withdrawal effect reflects a pharmacolo-

11

gical process.

Re-administration of caffeine reverses abstinence effectsAfter a period of abstinence during which the severity orincidence of a symptom or sign develops, the demonstra-tion that re-administration of caffeine rapidly and dosedependently reverses the abstinence effects helps toconfirm that the withdrawal effect reflects a pharmacolo-gical process.

Symptoms and signs of caffeine withdrawal

Symptoms of caffeine withdrawal

In this section, the withdrawal symptoms (i.e., categoriesof self-reported changes in mood or behavior) that wereassessed in the studies listed in Tables 1, 2, and 3, andsummarized in Table 4, are individually discussed andevaluated. For some symptoms, relevant case reports arealso discussed if they contribute to the evaluation. As inTable 4, the symptoms are sequenced from those that weredocumented in the greatest number of studies to those thatwere documented in the fewest number.

Of relevance to the assessment of each symptom isinformation about the “hit-rate” (i.e., the ratio of thenumber of times a symptom was found to be significantrelative to the number of times it was assessed). For thisanalysis, the number of studies in which each symptomwas documented was taken from the 57 experimentalstudies described in Tables 1 and 2 (and summarized incolumns 2–6 of Table 4). The number of studies in whicheach symptom was assessed was obtained from anevaluation of the same 57 experimental studies. In severalstudies (Comer et al. 1997; Schuh and Griffiths 1997;Garrett and Griffiths 1998; Jones et al. 2000) all of thesymptoms assessed could not be determined from thepublished article, and this information was obtained fromthe authors. [Methodological note: in the rare instance inwhich a study used a compound symptom descriptor (e.g.,alert/attentive/observant/able to concentrate) that couldpotentially apply to two symptom categories (e.g., alert/attentiveness and difficulty concentrating), the data werecounted in the single category that it best represented (e.g.,alert/attentiveness was counted as assessed and difficultyconcentrating was coded as not assessed).]

In interpreting the hit-rate, it should be noted that mostof the statistically significant effects reported are based ongroup mean data. Thus, a high hit-rate indicates that theabstinence condition is readily differentiated from thecomparison condition and usually reflects an intermediateto high incidence of the symptom. A low hit-rate indicatesthat the abstinence condition is not readily differentiatedfrom the comparison condition. However, it is important torecognize that a low hit-rate does not mean that a symptomis not valid because such a symptom may have a lowincidence and thus be undetected in a group statisticalanalysis. Furthermore, a symptom with a low hit-rate mayalso be clinically important to the extent that it may have

profound effects in a small percentage of the population(e.g., psychotic symptoms in the general population).Finally, a failure to detect a particular effect may reflectmethodological shortcomings of a study (e.g., insufficientperiod of abstinence, small sample size, insensitivemeasures).

Validity criteria For purposes of this review, the criteria forconcluding that a symptom is valid was the statisticaldemonstration of the symptom in six or more studies thatinclude two or more double-blind studies that usedmethodologies in which the conclusion of caffeine-with-drawal effects was not confounded by the direct effects ofcaffeine (i.e., Table 4, columns 4–6). As a conservativeapproach, two studies that used the same group of subjects(i.e., studies 5 and 6 in Table 4) were considered to be asingle study for purposes of judging validity.

Headache (descriptors: headache and headachy) Head-ache has been the most frequently assessed symptom (48experimental studies and 6 survey studies). Headache wasfound in 37 of 48 (77%) of the experimental studies inwhich it was assessed. The median percentage ofindividuals reporting headache in 19 experimental studieswas 47%, ranging from 9% to 100% across studies(Table 4). Of the 7 experimental studies that assessedheadache severity, the median percentage of subjectsreporting moderate to severe or maximum headache was50%. In 7 survey studies, the median percentage ofcaffeine users reporting caffeine-withdrawal headache was24%, ranging from 8% to 56%.

Headache as a caffeine-withdrawal symptom has beendemonstrated in studies comparing acute caffeine absti-nence with a preceding baseline condition, a caffeineadministration condition, and a chronic caffeine abstinencecondition (Table 4) and by comparing acute caffeineabstinence in caffeine consumers with non-consumers(Table 4). Studies have also shown that abstinence-induced headache is time-limited (Table 4) and is rapidly(usually within 30–60 min) and often completely reversedafter re-administration of caffeine (Driesbach and Pfeiffer1943; Goldstein et al. 1969; Roller 1981; Couturier et al.1997; Tinley et al. 2003), with the magnitude of reversalbeing an increasing function of the re-administeredcaffeine dose (Goldstein et al. 1969). Studies have alsoshown that the severity and incidence of headache afterabstinence were increasing functions of caffeine mainte-nance dose (Goldstein 1964; Silverman et al. 1992; Evansand Griffiths 1999) and duration of caffeine dosing (Evansand Griffiths 1999) before abstinence. Finally, whendifferent caffeine doses were substituted for the usualmaintenance dose, the severity and incidence of headacheincreased as the substituted dose decreased (Evans andGriffiths 1999). Caffeine-withdrawal headache has beencharacterized in these experimental studies as well as incase reports as being gradual in development (Driesbachand Pfeiffer 1943; Greden et al. 1980; Roller 1981;Griffiths et al. 1990), diffuse (Driesbach and Pfeiffer 1943;Greden 1974; Greden et al. 1980; Lader et al. 1996),

12

throbbing (Driesbach and Pfeiffer 1943; Greden 1974;Greden et al. 1980; Lader et al. 1996), severe (Griffithsand Woodson 1988; cf. Table 4), intensified with exerciseand Valsalva maneuver (Driesbach and Pfeiffer 1943), andphenomenologically distinct from migraine headache(Driesbach and Pfeiffer 1943).

Clinical reports, correlational analysis, and clusteranalysis of withdrawal symptoms have indicated thatnon-headache symptoms and signs of caffeine withdrawaldo not always co-vary with the presence of headache andcan occur in the absence of headache (Griffiths andWoodson 1988; Griffiths et al. 1990; Lader et al. 1996;Garrett and Griffiths 1998; Evans and Griffiths 1999). Thisindicates that non-headache symptoms represent distinctfeatures of the caffeine-withdrawal syndrome that canoccur independently of headache.

In summary, headache has been very frequently studied,has an intermediate incidence, and has been demonstratedunder a wide range of different methodological conditions.It is concluded that headache is a valid withdrawalsymptom.

Tiredness/fatigue (descriptors: tiredness, tired, fatigue,lazy, sluggish, lazy/sluggish/slow-moving, lethargic, leth-argy/fatigue/tired/sleepy, sluggish/slowed down, worn-out,and lassitude) Tiredness/fatigue was demonstrated in 32 of38 studies (84%). In 7 experimental studies providingincidence data (Table 4), the median percentage ofindividuals showing tiredness/fatigue was 27%. In 3survey studies, the percentage of subjects reportingtiredness/fatigue ranged between 21% and 56%. Tired-ness/fatigue has been demonstrated in studies comparingacute caffeine abstinence with a preceding baselinecondition, a caffeine administration condition, and achronic caffeine abstinence condition (Table 4), and bycomparing acute caffeine abstinence in caffeine consumerswith that in non-consumers (Table 4). Studies have alsoshown that abstinence-induced tiredness/fatigue is timelimited (Table 4), completely reversed after re-administra-tion of caffeine (Roller 1981), and an increasing functionof the duration of caffeine maintenance before abstinence(Evans and Griffiths 1999). Furthermore, when differentcaffeine doses are substituted for the usual maintenancedose, the magnitude of tiredness/fatigue increases as thesubstituted dose decreases (Evans and Griffiths 1999).Severity of tiredness/fatigue also appears to increase athigher daily caffeine maintenance doses (Rogers et al.1995).

In summary, tiredness/fatigue has been very frequentlystudied, has a low to moderate incidence, and has beendemonstrated under a wide range of methodologicalconditions. It is concluded that tiredness/fatigue is avalid withdrawal symptom.

Decreased energy/activeness (descriptors: decreased en-ergy, energetic, active, stimulated/active/energetic/excited,vigor, lively, stronger/more vigorous/more energy) De-creased energy/activeness was demonstrated in 24 of 32experimental studies (75%). In 4 experimental studies

providing incidence data (Table 4), the median percentageof individuals showing decreased energy/activeness was36%. Decreased energy/activeness has been demonstratedin studies comparing acute caffeine abstinence with apreceding baseline condition, a caffeine administrationcondition, and a chronic caffeine abstinence condition(Table 4), and by comparing acute caffeine abstinence incaffeine consumers with that in non-consumers (Table 4).Studies have have also shown that abstinence-induceddecreased energy/activeness is time limited (Table 4) andis rapidly (i.e., within 30–60 min) and completely reversedafter re-administration of caffeine (Goldstein et al. 1969).Studies have also shown that the severity of decreasedenergy/activeness is an increasing function of the caffeinemaintenance dose before abstinence (Rogers et al. 1995;Evans and Griffiths 1999).

In summary, decreased energy/activeness has been veryfrequently studied, has an intermediate incidence, and hasbeen demonstrated under a wide range of methodologicalconditions. It is concluded that decreased energy/active-ness is a valid withdrawal symptom.

Decreased alertness/attentiveness (descriptors: decreasedalertness, alert/attentive/observant, alert/attentive/able toconcentrate, and activity/alertness) Decreased alertness/attentiveness was demonstrated in 22 of 31 experimentalstudies (71%). Two experimental studies reported inci-dence data of 27% and 50% (Table 4). Decreasedalertness/attentiveness has been demonstrated in studiescomparing acute caffeine abstinence with a precedingbaseline condition, a caffeine administration condition,and a chronic caffeine abstinence condition (Table 4), andby comparing acute caffeine abstinence in caffeineconsumers with that in non-consumers (Table 4). Studieshave also shown that abstinence-induced decreasedalertness/attentiveness is time limited (Table 4) and israpidly (i.e., within 30–60 min) and completely reversedafter re-administration of caffeine (Goldstein et al. 1969),with the magnitude of reversal being an increasingfunction of the re-administered caffeine dose (Goldsteinet al. 1969). In addition, deficits in alertness/attentivenessare an increasing function of caffeine maintenance dosebefore abstinence (Rogers et al. 1995; Evans and Griffiths1999). Finally, when different caffeine doses are sub-stituted for the usual maintenance dose, the magnitude ofalertness/attentiveness decreases as the substituted dosedecreases (Evans and Griffiths 1999).

In summary, decreased alertness/attentiveness has beenvery frequently studied, may have an intermediateincidence, and has been demonstrated under a widerange of methodological conditions. It is concluded thatdecreased alertness/attentiveness is a valid withdrawalsymptom.

Drowsiness/sleepiness (descriptors: drowsiness, sleepy,drowsy/sleepy/tired, sedated, feel half awake, and de-creased wakefulness) Drowsiness/sleepiness was demon-strated in 21 of 27 experimental studies (78%). Across 5experimental studies providing incidence data (Table 4),

13

the median percentage of individuals showing drowsiness/sleepiness was 45%. In 3 survey studies, the percentage ofsubjects reporting tiredness/fatigue ranged between 18%and 59%. Drowsiness/sleepiness has been demonstrated instudies comparing acute caffeine abstinence with apreceding baseline condition, a caffeine administrationcondition, and a chronic caffeine abstinence condition(Table 4), and by comparing acute caffeine abstinence incaffeine consumers with that in non-consumers (Table 4).Studies have also shown that abstinence-induced drowsi-ness/sleepiness is time limited (Table 4) and is rapidly (i.e.,within 30–60 min) and completely reversed after re-administration of caffeine (Goldstein et al. 1969), with themagnitude of reversal being an increasing function of there-administered caffeine dose (Goldstein et al. 1969).Severity is positively correlated with daily caffeine dosebefore abstinence (Silverman et al. 1992).

In summary, drowsiness/sleepiness has been veryfrequently studied, has an intermediate incidence, andhas been demonstrated under a wide range of methodo-logical conditions. It is concluded that drowsiness/sleepiness is a valid withdrawal symptom.

Decreased contentedness/well-being (descriptors: de-creased contentedness, content/satisfied, well-being, daypositive, cheerful, happy, and increased miserable)Decreased contentedness/well-being was demonstrated in17 of 28 experimental studies (61%). In 2 experimentalstudies, 18% and 64% of participants showed decreases inmeasures of contentedness/well-being (Table 4). De-creased contentedness/well-being has been demonstratedin studies comparing acute caffeine abstinence with apreceding baseline condition, a caffeine administrationcondition, and a chronic caffeine abstinence condition(Table 4), and by comparing acute caffeine abstinence incaffeine consumers with that in non-consumers (Table 4).Studies have also shown that abstinence-induced de-creased contentedness/well-being is time limited (Table 4)and is rapidly (i.e., within 30–60 min) and completelyreversed after re-administration of caffeine (Goldstein etal. 1969), with the magnitude of reversal being anincreasing function of the re-administered caffeine dose(Goldstein et al. 1969). Severity also appears to increasewith higher daily maintenance caffeine doses beforeabstinence (Rogers et al. 1995).

In summary, decreased contentedness/well-being hasbeen very frequently studied, may have an intermediateincidence, and has been demonstrated under a wide rangeof methodological conditions. It is concluded thatdecreased contentedness/well-being is a valid withdrawalsymptom.

Decreased desire to socialize (descriptors: decreaseddesire to socialize, talkativeness, social disposition, andfriendliness) Decreased desire to socialize was demon-strated in 15 of 28 experimental studies (54%). Oneexperimental study reported incidence data of 79% fordecreased social disposition and a second study reported9% for decreased talkativeness (Table 4). Decreased desire

to socialize has been demonstrated in studies comparingacute caffeine abstinence with a preceding baselinecondition, a caffeine administration condition, and achronic caffeine abstinence condition (Table 4). Studieshave also shown that abstinence-induced decreased desireto socialize is time limited (Table 4) and that themagnitude is an increasing function of caffeine mainte-nance dose before abstinence (Evans and Griffiths 1999).

In summary, decreased desire to socialize has been veryfrequently studied with varied incidence. Although it hasbeen demonstrated under several different methodologicalconditions, it does not presently fulfill the criteria forvalidity.

Flu-like symptoms (descriptors: “flu-like symptoms” and“flu-like feelings”) An increase in this category wasdemonstrated in 9 of 17 (53%) experimental studies. Oneexperimental study reported incidence data of 31% for flu-like symptoms (Table 4). Flu-like symptoms as a symptomcategory have been demonstrated in studies comparingacute caffeine abstinence with a preceding baselinecondition, a caffeine administration condition, and achronic caffeine abstinence condition (Table 4). Studieshave also shown that abstinence-induced flu-like symp-toms are time-limited (Table 4) and that the severity of flu-like symptoms is an increasing function of caffeinemaintenance dose before abstinence (Lane and Phillips-Bute 1998; Evans and Griffiths 1999).

In summary, the flu-like symptoms category has beenfrequently studied, appears to have an intermediateincidence, and has been demonstrated under severalmethodological conditions. In addition to the foregoingstudies, one study (Swerdlow et al. 2000) showedcaffeine-abstinence-induced increases in a compositescale that included flu-like symptoms (i.e., feel sick,queasy, dizzy, and perspiring). It seems plausible thatendorsement of flu-like symptoms is related to aconstellation of somatic symptoms that include nausea/vomiting, muscle pain/stiffness, and heavy feelings inarms and legs (these individual symptoms are discussedbelow). Although most of the experimental studiesdemonstrating statistical increases in flu-like symptomshave involved comparisons of caffeine abstinence to acaffeine administration condition, the category appears toreflect a genuine withdrawal effect because endorsementof flu-like symptoms is time limited (Griffiths et al. 1990)and it is implausible that such placebo versus caffeinedifferences represent a direct effect of caffeine insuppressing naturally occurring flu-like symptoms. Thus,even though the flu-like symptoms category fails to meetour a priori criteria for validity, the category appears to bea valid caffeine-withdrawal effect. Furthermore, the flu-like symptoms category may be clinically importantbecause it may reflect significant distress.

Depressed mood (descriptors: depression, dejection, sad/depressed, and elevated scores on the Beck DepressionInventory) Symptoms of depressed mood were demon-strated in 9 of 29 experimental studies (31%). In 4

14

experimental studies providing incidence data, the medianpercentage of individuals reporting depressed mood was16% (range 11–36%; Table 4). In 2 survey studies, thepercentage of subjects reporting depression was 4% and9%. Depressed mood has been demonstrated in studiescomparing acute caffeine abstinence with a precedingbaseline condition, a caffeine administration condition,and a chronic caffeine abstinence condition (Table 4), andby comparing acute caffeine abstinence in caffeineconsumers with that in non-consumers (Table 4). Studiesalso showed that abstinence-induced depressed mood istime limited (Table 4).

In summary, depressed mood has been very frequentlystudied, has a low to moderate incidence, and has beendemonstrated under a wide range of methodologicalconditions. It is concluded that depressed mood is avalid withdrawal symptom.

Difficulty concentrating (descriptors: difficulty concen-trating, decreased concentration, and decreased ability toconcentrate) Difficulty concentrating was demonstrated in8 of 12 experimental studies (67%). One experimentalstudy reported incidence data of 79% and one surveystudy found that 11% of respondents reported difficultyconcentrating (Table 4). Difficulty concentrating has beendemonstrated in studies comparing acute caffeine absti-nence with a preceding baseline condition, a caffeineadministration condition, and a chronic caffeine abstinencecondition (Table 4). Studies have also shown that absti-nence-induced difficulty concentrating is time limited(Table 4), and the severity is positively correlated withdaily caffeine dose before abstinence (Lane 1997).

In summary, difficulty concentrating has been moder-ately frequently studied, may have a high incidence, andhas been demonstrated under a wide range of methodo-logical conditions. It is concluded that difficulty concen-trating is a valid withdrawal symptom.

Irritability (descriptors: irritability, irritable/cross/grum-py, and irritable/frustrated/angry/cross) Irritability wasdemonstrated in 8 of 23 experimental studies (35%). Twoexperimental studies reported incidence data of 29% and9% (Table 4). In 2 survey studies, the percentage ofsubjects reporting irritability during caffeine abstinencewas 21% and 20%. Irritability has been demonstrated instudies comparing acute caffeine abstinence with apreceding baseline condition, a caffeine administrationcondition, and a chronic caffeine abstinence condition(Table 4), and by comparing acute caffeine abstinence incaffeine consumers with that in non-consumers (Table 4).Studies have also shown that abstinence-induced irritabil-ity is time limited (Table 4) and is rapidly (i.e., within60 min) and completely reversed after re-administration ofcaffeine (Goldstein et al. 1969), with the magnitude ofreversal being an increasing function of the re-adminis-tered caffeine dose (Goldstein et al. 1969).

In summary, irritability has been very frequentlystudied, has a low to moderate incidence, and has beendemonstrated under a wide range of methodological

conditions. It is concluded that irritability is a validwithdrawal symptom.

Unmotivated for work (descriptors: decreased motivationfor work, urge to do tasks/work-related activities, andincreased unmotivated) Being unmotivated for work wasdemonstrated in 8 of 16 experimental studies (50%). Oneexperimental study reported incidence data of 57% fordecreased motivation for work and 64% for decreased urgeto do work (Table 4). In 2 survey studies, the percentage ofsubjects reporting being unmotivated for work was 16%and 18%. Unmotivated for work has been demonstrated instudies comparing acute caffeine abstinence with apreceding baseline condition and with a caffeine admin-istration condition (Table 4). Studies have also shown thatabstinence-induced “unmotivated for work” is time limited(Table 4), and the severity is positively correlated withdaily caffeine dose before abstinence (Lane 1997).

In summary, unmotivated for work has been frequentlystudied, has an intermediate incidence, and has beendemonstrated under several methodological conditions. Ithas also been described anecdotally (Driesbach andPfeiffer 1943). Although the data are suggestive, furtherresearch is needed to determine the validity of beingunmotivated for work as a withdrawal symptom.

Muzzy/foggy/not clearheaded (descriptors: muzzy/foggy/not clearheaded, muzziness, muddled, and decreasedclearheaded) Muzzy/foggy/not clearheaded was demon-strated in 7 of 18 experimental studies (39%). Oneexperimental study reported incidence data of 71%(Table 4). Muzzy/foggy/not clearheaded has been demon-strated in studies comparing acute caffeine abstinence witha preceding baseline condition, a caffeine administrationcondition, and a chronic caffeine abstinence condition(Table 4), and by comparing acute caffeine abstinence incaffeine consumers with that in non-consumers (Table 4).Studies have also shown that abstinence-induced muzzy/foggy/not clearheaded is time limited (Table 4), and theseverity is positively correlated with daily caffeine dosebefore abstinence (Lane 1997).

In summary, muzzy/foggy/not clearheaded has beenfrequently studied, may have a high incidence, and hasbeen demonstrated under a wide range of methodologicalconditions. It is concluded that muzzy/foggy/not clear-headed is a valid withdrawal symptom.

Yawning (no other descriptors) Self-reported yawning wasdemonstrated in 7 of 12 experimental studies (58%). Oneexperimental study reported incidence data of 43% andone survey study found that 21% of respondents reportedyawning (Table 4). Yawning has been demonstrated instudies comparing acute caffeine abstinence with apreceding baseline condition, a caffeine administrationcondition, and a chronic caffeine abstinence condition(Table 4). Studies have also shown that abstinence-induced yawning is time limited (Table 4), and theseverity is positively correlated with daily caffeine dosebefore abstinence (Silverman et al. 1992).

15

In summary, yawning has been moderately frequentlystudied, may have an intermediate incidence, and has beendemonstrated under several methodological conditions.Although the data are suggestive, further research isneeded to determine the validity of yawning as awithdrawal symptom.

Decreased self-confidence (descriptors: decreased self-confidence and confident) Decreased self-confidence wasfound in 7 of 17 experimental studies (41%). Oneexperimental study reported incidence data of 64%(Table 4). Decreased self-confidence has been demonstra-ted in studies comparing acute caffeine abstinence with apreceding baseline condition, a caffeine administrationcondition, and a chronic caffeine abstinence condition(Table 4). Studies have also shown that abstinence-induced decreased self-confidence is time limited(Table 4).

In summary, decreased self-confidence has beenfrequently studied and may have an intermediate inci-dence. Although it has been demonstrated under a severaldifferent methodological conditions, it does not presentlyfulfill criteria for validity.

Confusion–bewilderment [POMS] (no other descriptors)Confusion/bewilderment was found in 6 of 21 experi-mental studies (29%). Confusion–bewilderment has beendemonstrated in studies comparing acute caffeine absti-nence with a preceding baseline condition, a caffeineadministration condition, and a chronic caffeine abstinencecondition (Table 4). Studies have also shown that absti-nence-induced confusion–bewilderment is time limited(Table 4).

In summary, confusion–bewilderment has been veryfrequently studied and has been demonstrated underseveral methodological conditions. Although the data aresuggestive, it does not presently fulfill criteria for validity.

Total mood disturbance [POMS] (no other descriptors)Total mood disturbance has been shown in 6 of 12experimental studies (50%). Total mood disturbance hasbeen demonstrated in studies comparing acute caffeineabstinence with a preceding baseline condition, a caffeineadministration condition, and a chronic caffeine abstinencecondition (Table 4). Studies have also shown that theseverity of total mood disturbance is an increasingfunction of caffeine maintenance dose before abstinence(Evans and Griffiths 1999).

In summary, total mood disturbance has been moder-ately frequently studied and has been demonstrated underseveral methodological conditions. Although the data aresuggestive, further research is needed to determinevalidity.

Nausea/vomiting (descriptors: nausea/vomiting, nauseaand sickness, sick/nauseated/vomiting, vomiting, queasy,and upset stomach) Nausea/vomiting has been demon-strated in 6 of 24 experimental studies (25%). In 4experimental studies providing incidence data (Table 4),

the median percentage of individuals reporting nausea/vomiting was 21% (range 10–33%). In 2 survey studies,the percentage of subjects reporting nausea/vomitingduring caffeine abstinence was 3% and 21%. Nausea/vomiting has been demonstrated in studies comparingacute caffeine abstinence with a caffeine administrationcondition (Table 4).

In summary, nausea/vomiting has been very frequentlystudied and has a low to moderate incidence. Although inexperimental studies nausea/vomiting has only beendemonstrated statistically by comparing caffeine absti-nence with a caffeine administration condition, it appearsto be a genuine withdrawal symptom because it isimplausible that such placebo versus caffeine differencesrepresent a direct effect of caffeine in suppressingnaturally occurring nausea/vomiting. Instances of with-drawal-induced nausea/vomiting have been reported incase reports (Rainey 1985; Cacciatore et al. 1996),experimental studies (Griffiths et al. 1990; Silverman etal. 1992; Strain et al. 1994), and survey studies (Hughes etal. 1998; Oberstar et al. 2002). Thus, even though nausea/vomiting fails to meet our a priori criterion for validitybased on experimental studies alone, nausea/vomitingappears to be a valid caffeine-withdrawal symptom.Furthermore, it may be a clinically important symptombecause it is likely to reflect significant distress. It seemsplausible that this symptom is related to a constellation ofsymptoms that include other somatic complaints andendorsement of flu-like symptoms.

Muscle pain/stiffness (descriptors: muscle pain/stiffness,muscle joint ache, general muscle pains, and leg musclepains) Muscle pain/stiffness has been demonstrated in 4 of15 experimental studies (27%). The incidence of musclepain/stiffness was 43% in one study (Table 4). Musclepain/stiffness has been demonstrated in studies comparingacute caffeine abstinence with a preceding baselinecondition, a caffeine administration condition, and achronic caffeine abstinence condition (Table 4). Onestudy also showed that abstinence-induced muscle pain/stiffness is time limited (Table 4).

In summary, muscle pain/stiffness has been studiedfrequently, may have an intermediate incidence, and hasbeen demonstrated under a range of methodologicalconditions. The symptom of muscle pain/stiffness hasbeen described in case reports (Cobbs 1982; Stringer andWatson 1987) and is consistent with one report, whichinvolved a musculoskeletal examination during caffeineabstinence (Reeves et al. 1997). Although additionalstudies are needed to fulfill our a priori validity criteria, theconclusion that muscle pain/stiffness represents a truewithdrawal symptom unconfounded by the direct effectsof caffeine seems reasonable because it is improbable thatcaffeine suppresses naturally occurring muscle pain/stiffness.

16

Anxiety/nervousness (descriptors: anxiety, anxious, ner-vous, tense, and elevated scores on the State-Trait AnxietyInventory) Increased anxiety/nervousness has been dem-onstrated in 4 of 34 experimental studies (12%; Table 4)].However, decreased anxiety/nervousness has been dem-onstrated in 2 of the 36 studies (Griffiths et al. 1990;Comer et al. 1997). One experimental study reportedincidence data of 8% for increased anxiety (Table 4). In 3survey studies, the percentage of subjects reportingincreased anxiety during abstinence ranged between 10%and 29%. Anxiety/nervousness has been demonstrated instudies comparing acute caffeine abstinence with apreceding baseline condition and a caffeine administrationcondition (Table 4), and by comparing acute caffeineabstinence in caffeine consumers with that in non-consumers (Table 4).

Despite the fact that marked anxiety is proposed as acaffeine-withdrawal symptom in DSM-IV-TR (AmericanPsychiatric Association 2000), anxiety/nervousness doesnot fulfill the validity criteria. It is of note that 32experimental studies using a variety of methodologiesfailed to show increased anxiety/nervousness, none of thepositive studies used the most rigorous experimentaldesign involving a chronic abstinence condition, and 2experimental studies demonstrated significant decreases(Griffiths et al. 1990; Comer et al. 1997). However, it isalso noteworthy that anxiety/nervousness has beendescribed as a withdrawal symptom in case reports(Gibson 1981; Cobbs 1982; Rainey 1985; Adams et al.1993) and that anxiety/nervousness is also endorsed as awithdrawal symptom at low to moderate rates in threesurvey studies (Table 4). Although further research isneeded, it may be that the increased anxiety/nervousnessassociated with caffeine withdrawal in non-blind casereports and survey studies reflects increased anxiety inanticipation of experiencing unpleasant effects of caffeineabstinence.

Heavy feelings in arms and legs (no other descriptors)Heavy feelings in arms and legs were demonstrated in 4 of9 experimental studies (44%). Heavy feelings in arms andlegs have been demonstrated in studies comparing acutecaffeine abstinence with a preceding baseline condition, acaffeine administration condition, and a chronic caffeineabstinence condition (Table 4). Although the data aresuggestive, further research is needed to determine thevalidity of heavy feelings in arms and legs as a withdrawalsymptom.

Increased nighttime sleep duration/quality (descriptors:self-report ratings of sleep quality, duration, and onset tosleep) Self-reported increased nighttime sleep duration/quality was demonstrated in 4 of 4 experimental studies. Inthe two studies that assessed sleep duration, the increasewas about 30 min (Höfer and Bättig 1994b; James 1998).Increased nighttime sleep duration/quality has been dem-onstrated in studies comparing acute caffeine abstinencewith a preceding baseline condition and a caffeineadministration condition (Table 4).

Although increased nighttime sleep has been rarelystudied during caffeine abstinence, the magnitude ofincreased sleep duration is notable. The limited methodol-ogies in which it has been assessed do not permitdifferentiation between an effect of caffeine in decreasingsleep and a time limited caffeine-withdrawal effect.Further research is needed to determine whether this is avalid withdrawal symptom.

Analgesic use (no other descriptors) Self-reported anal-gesic use was demonstrated in 3 of 3 experimental studies.Two of these studies explicitly discouraged analgesic use(Silverman et al. 1992; Strain et al. 1994). Twoexperimental studies reported incidence data of 13% and45% (Table 4). Analgesic use has been demonstrated instudies comparing acute caffeine abstinence with apreceding baseline condition and a caffeine administrationcondition (Table 4). Studies also showed that abstinence-induced analgesic use was time limited (Table 4).

Although analgesic use during caffeine abstinence hasbeen rarely assessed, it is a potential indicator of theclinical significance of caffeine-withdrawal distress and isworthy of future study. Whether avoidance of caffeinewithdrawal contributes to chronic use of caffeine-contain-ing analgesics as suggested by some reports (Strain andGriffiths 1998; Bigal et al. 2002; but cf. Feinstein et al.2000) also merits further study.

Craving/strong desire to use (no other descriptors)Craving/strong desire to use was demonstrated in 2 of 2experimental studies. One experimental study reportedincidence data of 43% and one survey study reported that28% of respondents reported a strong desire to use(Table 4). Craving/strong desire to use has been demon-strated in studies comparing acute caffeine abstinence witha caffeine administration condition and has been shown tobe time limited (Table 4).

Although craving is often reported anecdotally duringcaffeine abstinence in the natural environment (Rippere1984; Gilbert 1986), it has been rarely assessed as acaffeine-withdrawal symptom. Further research is neededto determine its validity. Given its potential importance forunderstanding habitual caffeine consumption, a priorityshould be given to including measures of craving in futurecaffeine-withdrawal research.

Blurred vision (no other descriptors) Blurred vision wasdemonstrated in 2 of 11 experimental studies (18%). Theincidence of blurred vision was 14% in one experimentalstudy (Table 4). It has been demonstrated in studiescomparing acute caffeine abstinence with a precedingbaseline condition, a caffeine administration condition,and a chronic caffeine abstinence condition (Table 4). Onestudy also showed that abstinence-induced blurred visionwas time limited (Table 4).

In summary, blurred vision has been studied moderatelyfrequently, may have a low incidence, and has beendemonstrated under several methodological conditions.The symptom of blurred vision is also consistent with a

17

case report (Cobbs 1982). Further research is needed todetermine its validity.

Lightheaded/dizzy (no other descriptors) Lightheaded/dizzy was demonstrated in 3 of 18 experimental studies(17%). Lightheaded/dizzy has been demonstrated only instudies comparing acute caffeine abstinence with acaffeine administration condition.

Although lightheaded/dizzy has been frequently stu-died, it has only been demonstrated in three studies, one ofwhich was not blind. It is concluded that there is littleevidence supporting lightheaded/dizzy as a caffeine-with-drawal symptom.

Anger/hostility (descriptors: anger–hostility and angry)Anger/hostility was demonstrated in 2 of 20 experimentalstudies (10%). It has been demonstrated in studiescomparing acute caffeine abstinence with a caffeineadministration condition and a chronic caffeine abstinencecondition (Table 4), and by comparing acute caffeineabstinence in caffeine consumers with that in non-consumers (Table 4). Studies have also shown thatabstinence-induced anger/hostility is time limited(Table 4).

Although anger/hostility has been very frequentlystudied using various methodologies, it has only beendemonstrated in two studies, one of which was not blind.It is concluded that there is little evidence to suggest thatanger/hostility is a caffeine-withdrawal symptom.

Hot and cold spells (no other descriptors) This symptomwas demonstrated in 2 of 8 experimental studies (25%).Hot and cold spells have been demonstrated only instudies comparing acute caffeine abstinence with abaseline or a caffeine administration condition. Furtherresearch is needed to determine the validity.

Rhinorrhea (runny nose) [(descriptors: rhinorrhea andrunny nose)] Not shown in Table 4, rhinorrhea wasdemonstrated in one non-blind study (Roller 1981) of 12experimental studies that used a variety of methodologies.Caffeine-withdrawal-induced rhinorrhea was also de-scribed anecdotally in two reports (Driesbach and Pfeiffer1943; Greden et al. 1980). It is concluded that there is littleevidence to suggest that rhinorrhea is a caffeine-with-drawal symptom.

Diaphoresis (perspiration) [(descriptors: diaphoresis,perspiring, and sweating)] Not shown in Table 4, thissymptom was demonstrated in 2 of 19 experimentalstudies (11%). It is concluded that there is little evidence tosuggest that diaphoresis is a caffeine-withdrawal symp-tom.

Limb tremor (no other descriptors) Not shown in Table 4,this symptom was demonstrated in none of 10 experi-mental studies using a variety of methodologies. Theseobservations are consistent with seven studies that failed todocument objective measures of hand tremor during

caffeine abstinence (discussed below). It is concludedthat there is little evidence to suggest that self-reportedlimb tremor is a caffeine-withdrawal symptom.

Miscellaneous symptom categories for which there is noempirical support Several additional symptoms of poten-tial interest were evaluated in the studies described inTables 1 and 2 and were not significantly affected bycaffeine abstinence in any study. The symptoms, with thenumber of studies in which they were evaluated indicatedin parentheses, are heart pounding/palpitations (15),frequent urination (13), jittery/shaky (11), difficultysleeping (10), increased hunger/appetite (10), musclecramps (8), loss of sex drive (8), diarrhea (7), calm/relaxed (7), muscle twitches (5), irregular heartbeat (5),ringing in ears (5), thirsty (5), restless (4), shaky/weakness(3), constipation (2), chills (1), and numbing or tingling ofextremities (1).

Signs of caffeine withdrawal

In this section, withdrawal signs (i.e., objectively mea-sured behavioral or physiological effects) that wereassessed in the studies listed in Tables 1 and 2 areindividually discussed and evaluated.

Impaired behavioral and cognitive performance Thiscategory is comprised of tasks designed to assess variousaspects of performance impairment. Of 23 experimentalstudies that assessed performance during caffeine absti-nence, 11 (48%) reported significant impairment on one ormore measures. More specifically, impairment of tappingspeed occurred in 3 of 8 studies (Bruce et al. 1991;Silverman et al. 1992; Strain et al. 1994), impairment ofvisual vigilance occurred in 2 of 4 studies (Lane andPhillips-Bute 1998; Yeomans et al. 2002b), decreasedreaction time occurred in 2 of 7 studies (Rizzo et al. 1988;Robelin and Rogers 1998), impaired performance on adigit symbol substitution task occurred in 2 of 11 studies(Liguori and Hughes 1997; Liguori et al. 1997b), impairedperformance on a character recognition task occurred in 1study (James 1998), and impairment in 5 of 7 measures ina complex cognitive problem-solving task occurred in 1study (Streufert et al. 1995). Four studies assessingmemory found no evidence for impairment during caffeineabstinence. With regard to incidence, one experimentalstudy reported incidence data of 55% for decreasedtapping speed (Strain et al. 1994). Methodologically,impaired performance has been demonstrated in studiescomparing acute caffeine abstinence with a precedingbaseline condition (Silverman et al. 1992), a caffeineadministration condition (Silverman et al. 1992; Strain etal. 1994; Streufert et al. 1995; Liguori and Hughes 1997;Liguori et al. 1997b; James 1998; Lane and Phillips-Bute1998; Robelin and Rogers 1998; Yeomans et al. 2002b),and a chronic caffeine abstinence condition (Bruce et al.1991), and by comparing acute caffeine abstinence incaffeine consumers with that in non-consumers (Rizzo et

18

al. 1988). Studies have also shown that abstinence-inducedimpaired behavioral performance is time limited (Bruce etal. 1991).

In summary, impaired behavioral or cognitive perfor-mance has been very frequently studied, may have anintermediate incidence, and has been demonstrated under arange of methodological conditions. Furthermore, thetypes of impairments observed (i.e., psychomotor speed,vigilance, and cognitive performance) appear consistentwith the profile of validated withdrawal symptoms such astiredness/fatigue, decreased alertness, and difficulty con-centrating. It should be noted, however, that this categoryis comprised of heterogeneous measures of performanceimpairment and, at present, there is not enough informa-tion to reach a conclusion about the validity of any specificperformance measure. Future research should focus onmeasures that appear to be most sensitive to caffeineabstinence, including tapping, vigilance, reaction time, andcomplex cognitive problem solving.

Increased cerebral blood flow Increased cerebral bloodflow velocity has been demonstrated in 4 of 4 experi-mental studies (Mathew and Wilson 1985; Couturier et al.1997; Jones et al. 2000; Field et al. 2003; cf. Tables 1, 2).Increased cerebral blood flow has been demonstrated instudies comparing acute caffeine abstinence with apreceding baseline condition (Mathew and Wilson 1985;Couturier et al. 1997) and a caffeine administrationcondition (Jones et al. 2000; Field et al. 2003). Themagnitude of this effect is positively correlated with dailycaffeine dose before abstinence (Mathew and Wilson1985; Field et al. 2003).

The several studies indicating that caffeine abstinence isassociated with increases in cerebral blood flow are ofparticular interest because the effect may be related to avascular mechanism underlying the common withdrawalsymptom of headache (cf. Jones et al. 2000). However, themethodologies of the studies conducted to date do notpermit differentiation of the effects of caffeine per se oncerebral blood flow from the effects of caffeine withdraw-al. Further research is needed to determine the validity ofincreased cerebral blood flow as a withdrawal sign.

Changes in EEG Changes in quantitative electroenceph-alography (EEG) during caffeine abstinence was demon-strated in 3 of 3 experimental studies (Lader et al. 1996;Jones et al. 2000; Reeves et al. 2002; cf. Tables 1, 2).Effects included increases in EEG theta power (Jones et al.2000; Reeves et al. 2002), which have been associatedwith drowsiness. However, across studies, findings havebeen inconsistent across different EEG measures. Changesin EEG have been demonstrated in studies comparingacute caffeine abstinence with a preceding baselinecondition (Lader et al. 1996; Reeves et al. 2002) and acaffeine administration condition (Jones et al. 2000).

Although a few studies suggest that quantitative EEGmeasures might provide a physiological measure orcorrelate of caffeine withdrawal, the methodologies donot permit differentiation of the effects of caffeine per se

on EEG from the effects of caffeine withdrawal. Furtherresearch is needed to validate changes in EEG as awithdrawal sign.

Decreased blood pressure Of 11 experimental studies thatassessed blood pressure, three demonstrated decreases insystolic blood pressure (Streufert et al. 1995; Lader et al.1996; Phillips-Bute and Lane 1998) and one each showeddecreases in diastolic (Phillips-Bute and Lane 1998) andarterial (Lane 1997) blood pressure. The small magnitudeof this effect (about 6 mmHg for systolic blood pressure)suggests that, if this effect is time limited, it is notclinically important. Decreased blood pressure has beendemonstrated in studies comparing acute caffeine absti-nence with a preceding baseline condition (Lader et al.1996) and a caffeine administration condition (Streufert etal. 1995; Lane 1997; Phillips-Bute and Lane 1998). Onestudy showed the effect to be time limited (Lader et al.1996).

The studies cited above, as well as several studies thatexplicitly focused on cardiovascular outcome measures(Robertson et al. 1981; Ammon et al. 1983), do notprovide convincing evidence that caffeine abstinenceacutely decreases blood pressure.

Decreased motor activity Decreased motor activityassessed using electronic activity monitors was demon-strated in 2 of 2 experimental studies (Höfer and Bättig1994a,b; cf. Table 2). Decreased motor activity has beendemonstrated in studies comparing acute caffeine absti-nence with a preceding baseline condition (Höfer andBättig 1994a,b), and the effect was time limited (Höfer andBättig 1994a).

Although decreased motor activity has been only rarelystudied, it has face validity as a caffeine-withdrawal signbecause it plausibly could co-vary with the well-validatedsymptoms of increased tiredness/fatigue and decreasedenergy/activeness. Further research is needed to establishthe validity.

Skin conductance One experimental study assessed andfound a significant effect (decrease) in skin conductanceduring caffeine abstinence relative to a preceding baselinecondition (Lader et al. 1996; cf. Table 1). Further researchis needed to establish its validity.

Urinary epinephrine and norepinephrine Of two studiesthat assessed urinary epinephrine and norepinephrine(Robertson et al. 1981; Lane 1994; cf. Tables 1, 2), oneshowed decreased levels of epinephrine during caffeineabstinence relative to a caffeine condition (Lane 1994).The author interpreted this as an effect of caffeine ratherthan caffeine withdrawal. Further research is needed.

Heart rate Of 9 experimental studies that assessed heartrate, only one reported a significant effect (increase)during caffeine abstinence (Höfer and Bättig 1994a). Theeffect was neither time limited nor reversed by re-administration of caffeine (Höfer and Bättig 1994a). It is

19

concluded that there is no meaningful evidence thatcaffeine abstinence affects heart rate.

Hand tremor Of 7 experimental studies that assessedobjective measures of hand tremor, none showed asignificant effect of caffeine abstinence. These observa-tions are consistent with 10 studies that showed thatcaffeine abstinence did not affect subjective measures oflimb tremor (reviewed in Tables 1, 2, and discussedabove). It is concluded that there is little to suggest thatcaffeine abstinence affects hand tremor.

Incidence of clinically significant distress or impairmentin daily functioning

In individuals reporting caffeine-withdrawal symptoms,the severity can vary from mild to extreme. Clinicallysignificant distress and/or impairment of normal dailyactivities (e.g., work absence or going to bed because ofsymptoms) upon caffeine abstinence have been reported inclinical evaluations and experimental studies dating backover 170 years (Kingdon 1883; Bridge 1893; Driesbachand Pfeiffer 1943; Goldstein and Kaizer 1969; Greden etal. 1980; Rainey 1985; Smith 1987; Griffiths et al. 1990;Silverman et al. 1992; Adams et al. 1993; Weil and Rosen1993, p 187; Hampl et al. 1994; Strain et al. 1994;Cacciatore et al. 1996; Lader et al. 1996). Informationabout the proportion of regular caffeine consumers whoare at risk for experiencing clinically significant distressand/or functional impairment during abrupt caffeineabstinence is available from both prospective experimentalstudies and retrospective survey studies.

Prospective experimental studies showing clinicallysignificant distress or impairment

Six prospective experimental studies provide informationon the incidence of clinically significant distress orfunctional impairment during caffeine abstinence. Withregard to the experimental studies, in one double-blindstudy of 11 people who fulfilled criteria for DSM-IVsubstance dependence applied to caffeine use (medianintake 357 mg/day), 73% experienced significant disrup-tions in normal daily activities during a period ofexperimentally induced caffeine abstinence, includingleaving or missing work, making errors or costly mistakesat work, inability to care for children, and inability tocomplete school work (Strain et al. 1994). Five experi-mental studies have been conducted in normal subjectpopulations. One study, conducted in 22 medical andgraduate students who consumed 650–780 mg/dayexperimentally administered caffeine before abstinence,reported that headache “as extreme in severity as thesubjects had ever experienced” occurred in 55% of 38trials (Driesbach and Pfeiffer 1943). A double-blind studyof 62 caffeine consumers (mean intake 235 mg/day), whohad no knowledge that the study was about caffeine, found

that during caffeine abstinence 52% of subjects reportedmoderate to severe headache, and 8–11% showedabnormally high scores on standardized depression andfatigue scales (Silverman et al. 1992). In an open-endedinterview, several of these subjects also reported severefunctional impairment. In another double-blind study, 45%of 40 caffeine consumers (mean intake 360 mg/day)reported a “diffuse, throbbing headache,” with 28% ofthose also reporting “nausea and sickness” (Lader et al.1996). A third double-blind study evaluated the incidenceof functional impairment during abrupt caffeine abstinencein a group of 18 subjects (mean intake 231 mg/day) whoreported having had problems or symptoms whenpreviously stopping caffeine (Dews et al. 1999). Thestudy found that 39% of subjects spontaneously reportedcaffeine-withdrawal symptoms and 22% showed substan-tial decreases (≥1.5 points on a 4 point scale) in theirratings of daily functioning (e.g., at work and leisureactivities), although none of the subjects reported symp-toms judged to be “incapacitating” by the authors.Limitations of this study, which have been discussedelsewhere (Griffiths et al. 2003), include the relativelyunstructured and unmonitored conditions under whichdata were obtained and the relatively low caffeinemaintenance dose. Finally, one non-blind trial of caffeineabstinence reported 10% of 20 caffeine consumersreported nausea and vomiting (Couturier et al. 1997).

Overall, the incidence of clinically significant distress orfunctional impairment in prospective experimental studieswith normal subjects has varied from about 10% to as highas 55% (median 13%). The estimate of 13% is lower thanthe 73% rate of functional impairment shown in subjectswho met DSM-IV criteria for substance dependenceapplied to caffeine (Strain et al. 1994). The difference isnot due to differences in caffeine intake (the usualmaintenance dose in the dependent subjects was in therange of that of the studies in the normal subjects),suggesting that the caffeine-dependent subjects mayrepresent a subpopulation vulnerable to severe withdrawaleffects. It should be recognized that the 13% incidence ratefor normal subjects could underestimate the rate ofoccurrence in the general population. Although unstudied,it seems reasonable to suppose that there may be asubstantial subject selection bias in caffeine research in thegeneral population, with individuals who experienceclinically significant withdrawal being less likely tovolunteer for a study in which caffeine abstinence mightbe a possibility.

Retrospective survey studies showing clinicallysignificant distress or impairment

Four retrospective survey studies provide information onthe percentage of respondents reporting clinically signif-icant distress or functional impairment during caffeineabstinence. A survey study of 183 women showed that thepercentage endorsing that they would experience classiccaffeine-withdrawal symptoms (headache, lethargy, and

20

sleepiness) without their morning coffee increased as afunction of the number of cups of coffee consumed(Goldstein and Kaizer 1969). Consistent with functionalimpairment, endorsement of being “unable to workeffectively” also increased with cups consumed, with theeffect reported by 9% of all coffee consumers and 16% ofheavy consumers (>5 cups/day). In a population-basedrandom digit dial telephone survey study, caffeineconsumers were questioned about their experience withcaffeine withdrawal (Hughes et al. 1998). Of 71 whoreported having stopped or reduced caffeine for at least24 h during the past year, 11% reported experiencingheadache and other withdrawal symptoms that interferedwith their performance. This figure was 24% among thesubgroup who reported stopping caffeine in an attempt atpermanent abstinence. A third study asked a wide varietyof questions of people who telephoned to volunteer for apaid clinical research trial (Dews et al. 1999). Of 6,815daily caffeine consumers, 2.6% reported that problems orsymptoms during caffeine abstinence were severe enoughto interfere with normal activity (e.g., inability to concen-trate at work; lost time at work). The low rate ofendorsement of functional impairment in this study mayhave been due to a failure to exclude individuals who hadnot actually abstained from caffeine, as well as possibleunder reporting of symptoms because of a desire toparticipate in a paid research trial. A final retrospectivesurvey was conducted in 36 adolescent caffeine consumerswho endorsed two or more DSM-IV substance depen-dence criteria applied to caffeine use (Oberstar et al.2002). Twenty-one percent reported the caffeine-with-drawal symptom of “sick/nauseated/vomiting.”

Overall, the percentage of respondents reportingclinically significant distress or functional impairment inretrospective survey studies varied between 2.6% and 11%(median 9%) in general survey studies, and was 21% incaffeine-dependent adolescent subjects. The median esti-mate from the general survey studies is quite similar to theestimate from prospective experimental studies (9% and13%, respectively). As with the experimental studies, thereare several limitations inherent to survey studies that mayresult in underestimates. First, a portion of habitualcaffeine consumers may be unaware of caffeine-with-drawal symptoms because they never have had a period ofsustained abstinence. Furthermore, it has been demonstra-ted that as little as 25 mg/day of caffeine can prevent somewithdrawal symptoms (Evans and Griffiths 1999). Thus,small amounts of caffeine that are unknowingly consumedon days believed to be “caffeine-free” days may lead to anunderestimation of the incidence of clinically significantdistress during complete abstinence. Finally, it is possiblethat caffeine-withdrawal symptoms such as headache,fatigue, nausea, and muscle aches could be misattributedto other causes or ailments (e.g., common cold).

Expectancies and caffeine withdrawal

It has been speculated that knowledge and expectation arethe prime determinants of caffeine-withdrawal symptoms(Rubin and Smith 1999; Dews et al. 2002), leading to theconclusion that caffeine withdrawal is “controversial”(Dews et al. 1999; Rubin and Smith 1999). In drugresearch, the importance of expectancies has long beenacknowledged (Marlatt and Rohsenow 1980; Fillmore1994), and double-blind experimental methodologies wereexplicitly developed to help control for expectancy effects.However, what is known about withdrawal symptomsfrom abused drugs, including alcohol, opioids, sedatives,and cocaine, has been inferred from clinical observationsrather than documented in experimental studies in whichplacebo was substituted for active drug under blindconditions (Martin 1977; Weddington et al. 1990). Incontrast, the great majority (74%) of the 57 experimentalstudies of caffeine withdrawal outlined in Tables 1 and 2were conducted under double-blind conditions, with anadditional 11% conducted under single-blind conditions.Thus, caffeine-withdrawal research, in particular, seems todistinguish itself as having controlled for expectancyeffects better than most other research on drug withdrawal.

One approach to assessing the impact of expectancieson caffeine withdrawal is to compare the incidence ofcaffeine-withdrawal headache that has been reported indouble-blind and single-blind studies (studies 1–48 inTables 1 and 2) to that reported in non-blind studies(studies 49–55 in Table 2). The median incidence ofheadache (derived from column 7 of Table 4) for the blindand non-blind studies that reported such data is 45% and57%, respectively, suggesting a modest effect of expect-ancy.

However, it is important to recognize that the impact ofexpectancies is not entirely eliminated using double-blindmethods. Even in double-blind studies, expectancies canplay a role if subjects have knowledge of the purpose andconditions of the study. For instance, in extreme cases,subjects could be told that the purpose of a study is toinvestigate caffeine withdrawal and that they will bereceiving either caffeine or placebo. This informationcould functionally unblind subjects if they could dis-criminate between conditions based on immediate phar-macological effects or on early symptoms of withdrawal(Dews et al. 1999; Rubin and Smith 1999).

To address this issue, the blind studies in Tables 1 and 2were re-reviewed for information relevant to possibleexpectancy effects. This analysis revealed that many of theblind studies provided or probably provided sufficientinformation such that subjects could have been aware thatthe study involved caffeine or caffeine withdrawal.However, a substantial number of blind studies went tosome lengths to keep subjects uninformed about purposeor experimental conditions. For example, several studiesconducted at Johns Hopkins (Silverman et al. 1992; Schuhand Griffiths 1997; Garrett and Griffiths 1998; Jones et al.2000) instructed subjects that they could receive a varietyof compounds found in foods and beverages (e.g.,

21

chlorogenic acids, determines, caffeine, tannin, sugar,theophylline, or inactive placebo). To further divertattention away from caffeine, the dietary restrictionswere written without reference to caffeine and includedvarious foods and substances that do not contain caffeine(i.e., saccharin, aspartame [Nutrasweet], oysters, mussels,almonds, coconuts, poppy seeds, and all beverages exceptmilk, fruit juice, and water). Furthermore, as the researchunit at Johns Hopkins conducts a large number of studieswith a wide range of drugs other than caffeine, it isunlikely that conduct of these studies at that site wouldhave created subject expectations that the study involvedcaffeine. Two studies from other laboratories also usedinstructions and methods that reduced the likelihood ofexpectancy effects (Comer et al. 1997; Dews et al. 1999).Finally, several studies ruled out expectancy effects byusing debriefing procedures to explicitly determine whatsubjects understood or inferred about the experimentalconditions (van Dusseldorp and Katan 1990; Rogers et al.1995; James 1998; Robelin and Rogers 1998; Yeomans etal. 1998, 2002b; Tinley et al. 2003). From these studies inwhich expectancy effects are judged to be unlikely, themedian incidence of headache (derived from column 7 ofTable 4) for the studies that reported such data is 42%,which is similar to the incidence of headache for all blindstudies (45%). Furthermore, the profile of other symptomsdemonstrated in these studies in which expectancy effectswere unlikely is similar to those demonstrated in the largerset of studies.

In conclusion, while it is undoubtedly true that, as withthe assessment of other clinical phenomena, expectanciescould play some role in caffeine withdrawal, the evidencethat caffeine withdrawal is pharmacologically based isoverwhelming, and analysis of the published data does notsupport the hypothesis that expectancies are a primarydeterminant of caffeine-withdrawal symptoms. It shouldalso be recognized that, to the extent that expectanciesmay modestly enhance caffeine-withdrawal symptoms (assuggested from the comparison of blind and non-blindstudies described above), blind research studies mayactually underestimate the incidence and severity ofwithdrawal that occurs under more naturalistic clinicallyrelevant conditions. Future research should evaluate thispossibility using a methodology such as the balancedplacebo design, which experimentally manipulates bothcaffeine abstinence and expectancy of abstinence (Marlattand Rohsenow 1980; Juliano and Brandon 2002).

Time course of caffeine withdrawal

Numerous studies have shown that caffeine-withdrawalsymptoms typically emerge 12–24 h after abrupt caffeineabstinence (Driesbach and Pfeiffer 1943; Goldstein 1964;Goldstein et al. 1969; Griffiths et al. 1986, 1990;Richardson et al. 1995; Schuh and Griffiths 1997;Swerdlow et al. 2000; Tinley et al. 2003), which isconsistent with the short half-life of caffeine (4–6 h). Someevidence suggests that symptoms may emerge later

(>24 h) after abstinence from higher doses (e.g.,900 mg/day) of caffeine (Bruce et al. 1991; Evans andGriffiths 1992). In the few studies that provided detailedtime-course information for individual subjects, onset ofwithdrawal symptoms have been reported as early as 6 h(Roller 1981) and as late as 43 h (Griffiths et al. 1990)after abstinence. Caffeine-withdrawal symptoms havebeen shown to reach peak intensity between 20 h and51 h after abstinence (Griffiths et al. 1986, 1990; Evansand Griffiths 1992; Brauer et al. 1994; Höfer and Bättig1994a,b; Lader et al. 1996). The duration of caffeinewithdrawal has been shown to be 2–9 days (Griffiths et al.1986, 1990; van Dusseldorp and Katan 1990; Höfer andBättig 1994a), and the possibility of withdrawal headachesoccurring up to 21 days has been suggested (Richardson etal. 1995).

Parametric determinants of caffeine withdrawal

Chronic caffeine maintenance dose

There is good evidence that the incidence or severity ofcaffeine withdrawal increases with increases in the chronicdaily caffeine maintenance dose. The best evidence for thisrelationship comes from a prospective study that experi-mentally manipulated caffeine maintenance dose (100,300, and 600 mg/day) and showed monotonic increases inseveral withdrawal measures, with significantly greaterheadache and headache/poor mood demonstrated afterabstinence from 600 mg than 100 mg/day (Evans andGriffiths 1999). This study and a previous study (Griffithset al. 1990) also demonstrated that significant caffeinewithdrawal occurred after abstinence from a dose as low as100 mg/day.

The relationship between withdrawal incidence orseverity and usual caffeine dose has also been demonstra-ted in studies of self-reported caffeine maintenance dose,including retrospective survey studies (Goldstein andKaizer 1969), experimental studies (Goldstein 1964;Silverman et al. 1992; Rogers et al. 1995; Lader et al.1996; Lane 1997; Lane and Phillips-Bute 1998), andstudies of post-operative headache (Galletly et al. 1989;Fennelly et al. 1991). It should be noted, however, that therelationship between withdrawal and self-reported caffeineintake appears relatively weak because it has not beendemonstrated in some studies (Verhoeff and Millar 1990;Hughes et al. 1993; Höfer and Bättig 1994a), andsignificant correlations between withdrawal measuresand caffeine intake are low and inconsistent acrossdifferent withdrawal measures (Silverman et al. 1992;Lane 1997; Lane and Phillips-Bute 1998).

Acute decreases in caffeine maintenance dose

When lower caffeine doses are substituted for the usualmaintenance dose, withdrawal severity increases as thesubstituted dose decreases. One study maintained indivi-

22

duals on 300 mg caffeine/day and then substituted a rangeof lower doses (Evans and Griffiths 1999). The studyshowed that a substantial reduction in caffeine dose (to≤100 mg/day) was necessary to produce caffeine with-drawal and that even a dose of 25 mg/day was sufficient toavoid significant caffeine-withdrawal headache. Thisresult suggests that a substantial percentage reduction incaffeine consumption is necessary to manifest the fullcaffeine-withdrawal syndrome.

Duration of caffeine maintenance

Three studies demonstrated that caffeine withdrawal canoccur after a relatively short duration of caffeine mainte-nance (Driesbach and Pfeiffer 1943; Griffiths et al. 1986;Evans and Griffiths 1999). One of these studies (Evansand Griffiths 1999), which was conducted in caffeineconsumers who were caffeine abstinent for 7 days beforethe period of caffeine maintenance, showed no withdrawaleffects after a single day of exposure to caffeine (300 mg/day). However, significant withdrawal symptoms occurredafter 3 consecutive days of caffeine, with somewhatgreater severity demonstrated after 7 and 14 consecutivedays of exposure. Another study (Driesbach and Pfeiffer1943) showed that caffeine-withdrawal headache occurredin three individuals who normally totally abstained fromcaffeinated beverages, but who were given increasingdoses of caffeine over 6 days or 7 days up to 650–780 mg/day.

Within-day frequency of dosing during caffeinemaintenance

Most studies of caffeine withdrawal have involvedcaffeine abstinence after a maintenance period thatinvolved multiple caffeine doses each day. The onlystudy to vary the within-day frequency of caffeine dosingshowed that the range and severity of caffeine-withdrawalsymptoms was similar when caffeine maintenance in-volved 300 mg taken as single dose in the morningcompared with 100 mg taken at three time points acrossthe day (Evans and Griffiths 1999). This indicates thatonce-a-day dosing with caffeine is sufficient for producingwithdrawal symptoms.

Re-administration of caffeine reverses abstinenceeffects

After a period of abstinence during which the severity orincidence of a symptom or sign develops, re-administra-tion of caffeine rapidly (usually within 30–60 min) andoften completely reverses withdrawal (Driesbach andPfeiffer 1943; Goldstein et al. 1969; Roller 1981;Couturier et al. 1997; Tinley et al. 2003), with themagnitude of reversal being an increasing function of there-administered caffeine dose (Goldstein et al. 1969).

Individual differences in caffeine withdrawal

There are differences within and across individuals withrespect to the incidence of caffeine withdrawal. Asdiscussed above and shown in Table 4, only about 50%of subjects in experimental studies report headache afterany single occasion of caffeine abstinence, and theseverity of headache can vary from mild to extreme.One study that examined six repeated blind abstinencetrials in seven subjects maintained on 100 mg/day ofcaffeine documented differences within and across sub-jects: one subject never showed substantial headache,some subjects showed consistent headaches, while othersreported headaches on some trials but not others (Griffithset al. 1990). A second study that analyzed the effects of sixrepeated abstinence trials showed that at least 36% ofsubjects, who showed statistically significant elevations inheadache, failed to report this effect consistently acrossrepeated trials (Hughes et al. 1993). There is evidence thatgenetic factors may play a role in some differences amongindividuals. One study of 1,934 female twins found thatthere was significantly greater concordance of DSM-IV-defined caffeine withdrawal among monozygotic twins(41%) than dizygotic twins (18%), yielding an estimatedbroad heritability of 35% (Kendler and Prescott 1999).

Other than the role of chronic maintenance dose(discussed previously), very little is known about thedeterminants of individual differences in caffeine with-drawal. The results of one study suggested that individualswho eliminated caffeine slowly were less likely toexperience sedation during withdrawal (Lader et al.1996). Whether females, individuals with histories ofprevious drug dependence including cigarette smoking, orindividuals with polymorphisms in the A1 and A2A

adenosine receptor genes are at greater risk of caffeinewithdrawal is worthy of future research (Strain et al. 1994;Dews et al. 1999; Alsene et al. 2003).

The role of caffeine withdrawal in the habitualconsumption of caffeine

Research indicates that avoidance of abstinence-associatedwithdrawal symptoms plays a central role in the habitualconsumption of caffeine. This relationship has been shownin retrospective questionnaire studies (Goldstein andKaizer 1969) and in double-blind experimental studiesthat assessed direct behavioral measures of caffeinereinforcement or preference (Griffiths et al. 1986; Hugheset al. 1993; Liguori and Hughes 1997; Schuh and Griffiths1997; Garrett and Griffiths 1998) and beverage flavorpreferences (Rogers et al. 1995; Yeomans et al. 1998,2000a, 2001, 2002a; Tinley et al. 2003).

One study of caffeine reinforcement, for example,showed that moderate caffeine consumers who reportedcaffeine-withdrawal symptoms (i.e., headache, drowsi-ness) after drinking decaffeinated coffee were more thantwice as likely to choose caffeinated over decaffeinated

23

coffee in choice tests (Hughes et al. 1993). In studies thatprospectively manipulated caffeine physical dependence,subjects chose caffeine more than twice as often when theywere physically dependent than when they were notphysically dependent (Griffiths et al. 1986; Garrett andGriffiths 1998).

Withdrawal also plays an important role in the devel-opment of preferences for flavors paired with caffeine. Inthese studies, caffeine consumers who abstained fromcaffeine overnight and were repeatedly exposed to a novelflavored drink paired with caffeine showed increasedratings of drink pleasantness or preference compared withcaffeine consumers who received placebo-paired drinks(Rogers et al. 1995; Yeomans et al. 1998). It has beendemonstrated that the development of such flavorpreference requires that subjects be caffeine deprived attraining and testing (Yeomans et al. 1998, 2000b, 2001,2002a). Furthermore, the effects are not observed incaffeine non-consumers or in long-term abstinent con-sumers (Rogers et al. 1995; Tinley et al. 2003). It seemslikely that, in the natural environment, withdrawal-depen-dent conditioned flavor preferences play an important rolein development of strong consumer preferences forspecific kinds of caffeine-containing beverages.

DSM-IV-TR and ICD-10 diagnostic criteria for caffeinewithdrawal

The potential for caffeine withdrawal to cause clinicallysignificant distress or impairment in functioning isreflected by the inclusion of caffeine withdrawal as anofficial diagnosis in ICD-10 (World Health Organization1992a,b) and as a proposed research diagnosis in DSM-IV-TR (American Psychiatric Association 2000). Caffeinewithdrawal was included in DSM-IV as a proposeddiagnosis rather than an official diagnosis to encouragefurther research on the range and specificity of caffeine-withdrawal symptoms (Hughes 1994). The DSM-IV-TRproposed research criteria for withdrawal are: (A)prolonged daily use of caffeine; (B) abrupt cessation ofcaffeine use or reduction in the amount of caffeine used,closely followed by headache and one (or more) of thefollowing symptoms: (1) marked fatigue or drowsiness,(2) marked anxiety or depression, (3) nausea or vomiting;(C) the symptoms in criterion B cause clinically significantdistress or impairment in social, occupational, or otherimportant areas of functioning; (D) the symptoms are notdue to the direct physiological effects of a general medicalcondition (e.g., migraine, viral illness) and are not betteraccounted for by another mental disorder.

The official ICD-10 diagnosis for withdrawal fromcaffeine does not specify specific symptoms for makingthe diagnosis (World Health Organization 1992a,b). TheICD-10 diagnostic criteria for research includes thediagnosis of a “withdrawal state from other stimulants,including caffeine” (World Health Organization 1993).The criteria involve: (A) cessation or reduction of caffeineuse after prolonged use; (B) dysphoric mood (for instance,

sadness or anhedonia); and (C) two or more of thefollowing: (1) lethargy and fatigue, (2) psychomotorretardation or agitation, (3) craving for stimulant drugs,(4) increased appetite, (5) insomnia or hypersomnia, (6)bizarre or unpleasant dreams.

Proposed revision of DSM-IV-TR and ICD-10diagnostic criteria for caffeine withdrawal

The great majority of the research literature on caffeinewithdrawal has been published since the DSM-IV WorkGroup (in 1994) and the World Health Organization (in1993) formulated their respective criteria for proposedresearch diagnoses (cf. Tables 1, 2, 3). This sectionfocuses on a proposed revision of the DSM, althoughmany of the considerations would be applicable to arevision of ICD-10. The research summarized in thisreview further documents that caffeine withdrawal may besevere enough to warrant clinical attention, which wasquestioned in DSM-III-R (American Psychiatric Associa-tion 1987). Moreover, the research literature now providesa sound empirical basis for updating the criteria forcaffeine withdrawal. Problems with the DSM-IV-TRcriteria are that they (1) do not reflect the apparentindependence of headache and non-headache withdrawalsymptoms (cf. section on headache); (2) do not includesymptoms that have now been well-validated; and (3)include a symptom (i.e., anxiety) for which there is littleempirical support.

Based on the current evaluation of the researchliterature, we now propose that part B of the DSMresearch criteria be changed to indicate that a caffeine-withdrawal diagnosis requires: abrupt cessation of caffeineuse or reduction in the amount of caffeine used, closelyfollowed by three or more of the following: (1) headache,(2) fatigue or drowsiness, (3) dysphoric mood, depressedmood, or irritability, (4) difficulty concentrating, and (5)flu-like somatic symptoms, nausea, vomiting, or musclepain/stiffness.

The proposed five clusters of symptoms are based onthe 13 symptoms from this review that are the bestcandidates for describing the caffeine-withdrawal syn-drome. The new criteria address the 1994 DSM-IV WorkGroup concerns that there were too few validated symp-toms and there may be overlap between fatigue anddrowsiness (Hughes 1994). The decision to propose thediagnosis based on three of the five clusters is aconservative strategy to prevent over-diagnosis of thedisorder, which was also a concern of the DSM-IV WorkGroup (Hughes 1994). Although some of the proposedsymptoms have high prevalence and other etiologies, thisis also true of other withdrawal diagnoses recognized byDSM (e.g., cocaine withdrawal, nicotine withdrawal,opioid withdrawal).

It should be recognized that although the individualsymptoms have been empirically validated, the symptomclusters are conceptually rather than empirically derived. Itwould be informative if future research with a suitably

24

large group of individuals assessed all of the symptomsvalidated in this review and used statistical procedures toempirically differentiate among clusters of symptoms(e.g., cluster analysis or multiple regression), as well asdetermine their ability to predict clinically significantdistress.

Summary/conclusions

The present paper represents the most comprehensivereview and analysis of the effects of caffeine abstinence inhumans published to date. The purpose of this analysiswas to empirically validate specific symptoms and signs,and to appraise important features of the caffeine-withdrawal syndrome.

Of 49 symptom categories and 9 sign categoriesidentified from 57 experimental and 9 retrospective surveystudies, the following 10 symptom categories fulfilledmethodologically rigorous validity criteria: headache,tiredness/fatigue, decreased energy/activeness, decreasedalertness/attentiveness, drowsiness/sleepiness, decreasedcontentedness/well-being, depressed mood, difficulty con-centrating, irritability, and muzzy/foggy/not clearheaded.In addition, flu-like symptoms, nausea/vomiting, andmuscle pain/stiffness were judged likely to representvalid symptom categories. The percentage of subjectsreporting headache was 50% in experimental studies and24% in retrospective survey studies. The percentage ofsubjects reporting clinically significant distress or func-tional impairment was 13% in prospective experimentalstudies and 9% in retrospective survey studies.

Data supporting the following 13 symptom and signcategories were judged as suggestive and in need offurther research: decreased desire to socialize, unmoti-vated for work, decreased self-confidence, total mooddisturbance (POMS), yawning, heavy feelings in arms andlegs, increased nighttime sleep quality/duration, analgesicuse, craving/strong desire to use, impaired behavioral andcognitive performance (objectively measured), decreasedmotor activity (objectively measured), increased cerebralblood flow, and EEG changes.

Onset of withdrawal symptoms typically occurs 12–24 h after abstinence, with peak intensity occurring at 20–51 h, and duration of withdrawal ranging between 2 daysand 9 days. Re-administration of caffeine rapidly and oftencompletely reverses withdrawal. The incidence or severityof symptoms increases with increases in the chronic dailymaintenance dose. Symptoms may occur upon abstinencefrom chronic caffeine exposure at doses as low as 100 mg/day, and upon abstinence following only 3–7 days ofcaffeine exposure at higher doses. There is good evidencethat avoidance of withdrawal symptoms plays a centralrole in the habitual consumption of caffeine by increasingthe reinforcing effects of caffeine and preference for tastespaired with caffeine. Overall, the evidence that caffeinewithdrawal is pharmacologically based is overwhelming,and analysis of the published data does not support the

hypothesis that expectancies are a primary determinant ofcaffeine-withdrawal symptoms.

In addition to the previously discussed researchpriorities concerning the DSM-IV diagnosis of caffeinewithdrawal, future studies should determine the validity ofthe symptom and sign categories identified above as inneed of further research, and investigate how vulnerabilityto caffeine withdrawal is affected by gender, drug abusehistories, caffeine metabolism, genetics, behavioral con-ditioning, personality, and other factors. It would bevaluable if future reports of research on caffeine with-drawal provided individual subject data in addition to theusual group data to provide further information aboutindividual differences in severity of the withdrawalsyndrome. Future research should provide more detailedinformation about the time course (i.e., onset, duration,and offset) of individual withdrawal symptoms. Further-more, the impact of gradual reduction of caffeine onwithdrawal symptoms should be characterized. Althoughtolerance and withdrawal are thought to be functionallyrelated phenomena, no research has investigated therelationship between the extent of caffeine tolerance andmagnitude of withdrawal. Finally given the high rate ofcaffeine use in children, caffeine withdrawal in childrendeserves much more systematic study (Goldstein andWallace 1997; Bernstein et al. 1998).

In conclusion, although descriptions in the medicalliterature of the caffeine-withdrawal syndrome date backmore than 170 years, a solid empirical parametric analysisof the phenomenon has only begun to emerge in recentyears. Arguably, the caffeine-withdrawal syndrome hasbeen more rigorously and completely characterized thanwithdrawal from any other drug and can now serve as amodel system for evaluating drug withdrawal phenomena.Despite this progress, we can anticipate that futureresearch with caffeine will provide further valuableinsights into the world’s most widely consumed mood-altering drug, as well as insights into drug withdrawaleffects more generally.

Acknowledgements Preparation of this manuscript was supported,in part, by National Institute on Drug Abuse grant R01 DA03890.Roland Griffiths has been a consultant to pharmaceutical companies,International Food Information Counsel, International Life SciencesInstitute, and the legal profession on issues related to caffeineeffects, withdrawal, and dependence. The authors thank KristenMcCausland and Kimberly Mudd for their assistance with datamanagement.

25

Appendix A—Studies that did not statistically documentsymptom or sign

Symptom or sign Studiesa

Alertness/attentiveness 10, 14, 16, 22, 23, 28, 38, 46, 55Anger/hostility 4, 6, 9, 10, 15, 18, 22, 25, 26, 27, 28, 33,

34, 35, 36, 37, 46, 55Anxiety/nervousness 4, 5, 6, 10, 14, 15, 16, 17, 18, 19, 20, 21,

23, 24, 25, 26, 27, 28, 29, 33, 34, 35, 36,37, 41, 44, 45, 46, 55, 56

Blood pressure 3, 13, 38, 43, 44, 45, 54Blurred vision 5, 19, 21, 25, 26, 37, 46, 55, 56Calm/relaxedb 16, 17, 18, 21, 27, 31, 39Chills 21Circular lights taskb 4Confusion–bewilder-ment

4, 5, 10, 15, 17, 18, 21, 25, 28, 33, 35, 36,37, 46, 55

Constipationb 5, 6Contentedness/well-being

4, 10, 14, 15, 20, 27, 28, 29, 37, 39, 46

Depression 2, 4, 10, 18, 19, 20, 21, 22, 23, 24, 25, 28,29, 33, 34, 35, 36, 37, 46, 55

Desire to socialize 15, 18, 19, 21, 23, 24, 25, 26, 27, 28, 29,35, 36

Diaphoresis 5, 6, 9, 14, 15, 19, 20, 21, 22, 23, 24, 25,26, 37, 46, 55, 56

Diarrheab 5, 6, 14, 15, 22, 23, 24Difficulty sleeping/insomniab

9, 14, 15, 19, 21, 22, 23, 24, 46, 55

Difficulty concentrating 21, 25, 28, 46Digit symbol substitu-tion task

9, 14, 15, 20, 21, 22, 28, 37, 55

Divided attentionb 21Drowsiness/sleepiness 5, 10, 18, 22, 25, 37Energy/activeness 10, 14, 16, 17, 25, 36, 42, 46Flu-like symptoms 19, 21, 26, 28, 35, 37, 46, 56Frequent urinationb 9, 14, 15, 19, 22, 23, 24, 25, 26, 37, 46,

55, 56Grammatical/logicalreasoningb

28, 55

Hand tremor (objec-tively measured)b

14, 15, 17, 20, 22, 23, 24

Headache 14, 19, 22, 24, 26, 28, 29, 31, 37, 40, 41Heart pounding/palpita-tions

9, 14, 15, 19, 20, 21, 22, 23, 24, 25, 26,37, 46, 55, 56

Heart rate 3, 13, 19, 38, 43, 45, 54, 55Heavy feelings in armsand legs

19, 21, 25, 26, 56

Hot and cold spells 19, 25, 26, 37, 46, 55Hunger/appetiteb 17, 18, 20, 21, 22, 23, 24, 29, 31, 42Irregular heartbeat 14, 15, 22, 23, 24Irritability 4, 10, 14, 15, 16, 20, 21, 23, 24, 25, 28,

29, 37, 46, 55Jittery/shakyb 5, 6, 9, 28, 31, 37, 40, 41, 42, 46, 55Lightheaded/dizzy 5, 9, 14, 15, 19, 20, 21, 22, 23, 24, 25, 26,

37, 46, 56Limb tremorb 5, 6, 9, 19, 25, 26, 37, 46, 55, 56Loss of sex driveb 9, 19, 25, 26, 37, 46, 55, 56Memory/recallb 9, 21, 28, 37Muscle crampsb 9, 19, 25, 26, 37, 46, 55, 56Muscle pain/stiffness 9, 19, 20, 21, 25, 26, 37, 44, 46, 55, 56Muscle twitchesb 14, 15, 22, 23, 24Muzzy/foggy/not clear-headed

5, 8, 19, 25, 26, 28, 29, 31, 37, 42, 46

Nausea/vomiting/upsetstomach

2, 4, 5, 6, 9, 10, 14, 15, 19, 20, 21, 25, 26,37, 45, 46, 55, 56

Numbing or tingling inextremities

21

Symptom or sign Studiesa

Numerical Stroop taskb 9Problem solving tasks 19Reaction time 9, 17, 28, 41, 55Restlessb 21, 22, 23, 24Rhinorrhea 5, 6, 9, 19, 21, 25, 26, 37, 46, 55, 56Ringing in earsb 14, 15, 22, 23, 24Self-confidence 17, 18, 19, 22, 25, 26, 29, 37, 55, 56Shaky/weaknessb 22, 23, 24Symbol copy testb 20Tapping speed 17, 20, 28, 29, 55Thirst/thirstyb 17, 18, 29, 31, 42Tiredness/fatigue 10, 22, 27, 31, 37, 42Total mood disturbance 10, 25, 33, 37, 41, 46Unmotivated for work 14, 19, 25, 26, 28, 37, 46, 55Urinary epinephrine ornorepinephrine

3

Yawning 5, 25, 26, 37, 46Visual vigilance 20, 21aNumbers in table refer to the entry number in Tables 1 and2bSymptom or sign was not documented in any study

References

Adams D, Ditzler T, Haning WF (1993) Primary caffeine depen-dence: a case report. Hawaii Med J 52:190–191 (see also page194)

Alsene K, Deckert J, Sand P, de Wit H (2003) Association betweenA2a receptor gene polymorphisms and caffeine-induced anx-iety. Neuropsychopharmacology 28:1694–1702

American Psychiatric Association (1987) Diagnostic and statisticalmanual of mental disorders, 3rd edn. American PsychiatricAssociation, Washington (revised)

American Psychiatric Association (1994) Diagnostic and statisticalmanual of mental disorders, 4th edn. American PsychiatricAssociation, Washington

American Psychiatric Association (2000) Diagnostic and statisticalmanual of mental disorders, 4th edn. American PsychiatricAssociation, Washington (text revision)

Ammon HP, Bieck PR, Mandalaz D, Verspohl EJ (1983) Adaptationof blood pressure to continuous heavy coffee drinking in youngvolunteers: a double-blind crossover study. Br J Clin Pharmacol15:701–706

Barone JJ, Roberts HR (1996) Caffeine consumption. Food ChemToxicol 34:119–129

Bernstein GA, Carroll ME, Dean NW, Crosby RD, Perwien AR,Benowitz NL (1998) Caffeine withdrawal in normal school-agechildren. J Am Acad Child Adolesc Psychiatry 37:858–865

Bernstein GA, Carroll ME, Thuras PD, Cosgrove KP, Roth ME(2002) Caffeine dependence in teenagers. Drug AlcoholDepend 66:1–6

Bigal ME, Sheftell FD, Rapoport AM, Tepper SJ, Lipton RB (2002)Chronic daily headache: identification of factors associatedwith induction and transformation. Headache 42:575–581

Brauer LH, Buican B, de Wit H (1994) Effects of caffeinedeprivation on taste and mood. Behav Pharmacol 5:111–118

Brice CF, Smith AP (2002) Effects of caffeine on mood andperformance: a study of realistic consumption. Psychopharma-cology (Berl) 164:188–192

Bridge N (1893) Coffee-drinking as a frequent cause of disease.Trans Assoc Am Physicians 8:281–288

Bruce M, Scott N, Lader M, Marks V (1986) The psychopharma-cological and electrophysiological effects of single doses ofcaffeine in healthy human subjects. Br J Clin Pharmacol 22:81–87

26

Bruce M, Scott N, Shine P, Lader M (1991) Caffeine withdrawal: acontrast of withdrawal symptoms in normal subjects who haveabstained from caffeine for 24 hours and for 7 days. JPsychopharmacol 5:129–134

Cacciatore R, Helbling A, Jost C, Hess B (1996) Episodic headache,diminished performance and depressive mood. SchweizRundsch Med Prax 85:727–729

Cobbs LW (1982) Lethargy, anxiety, and impotence in a diabetic.Hosp Pract (Off Ed) 17:67 (see also pages 70 and 73)

Comer SD, Haney M, Foltin RW, Fischman MW (1997) Effects ofcaffeine withdrawal on humans living in a residential labora-tory. Exp Clin Psychopharmacol 5:399–403

Couturier EG, Laman DM, van Duijn MA, van Duijn H (1997)Influence of caffeine and caffeine withdrawal on headache andcerebral blood flow velocities. Cephalalgia 17:188–190

Denaro CP, Benowitz NL (1991) Caffeine metabolism: dispositionin liver disease and hepatic-function testing. In: Watson RR(ed) Drug and alcohol abuse reviews. Liver pathology andalcohol, vol 2. The Human, Totowa, pp 513–539

Dews PB, Curtis GL, Hanford KJ, O’Brien CP (1999) Thefrequency of caffeine withdrawal in a population-based surveyand in a controlled, blinded pilot experiment. J Clin Pharmacol39:1221–1232

Dews PB, O’Brien CP, Bergman J (2002) Caffeine: behavioraleffects of withdrawal and related issues. Food Chem Toxicol40:1257–1261

Driesbach RH, Pfeiffer C (1943) Caffeine-withdrawal headache. JLab Clin Med 28:1212–1219

Dusseldorp M van, Katan MB (1990) Headache caused by caffeinewithdrawal among moderate coffee drinkers switched fromordinary to decaffeinated coffee: a 12 week double blind trial.BMJ 300:1558–1559

Edelstein BA, Keaton-Brasted C, Burg MM (1983) The effects ofcaffeine withdrawal on cardiovascular and gastrointestinalresponses. Health Psychol 2:343–352

Evans SM, Griffiths RR (1992) Caffeine tolerance and choice inhumans. Psychopharmacology (Berl) 108:51–59

Evans SM, Griffiths RR (1999) Caffeine withdrawal: a parametricanalysis of caffeine dosing conditions. J Pharmacol Exp Ther289:285–294

Evans SM, Critchfield TS, Griffiths RR (1994) Caffeine reinforce-ment demonstrated in a majority of moderate caffeine users.Behav Pharmacol 5:231–238

Feinstein AR, Heinemann LA, Dalessio D, Fox JM, Goldstein J,Haag G, Ladewig D, O’Brien CP (2000) Do caffeine-contain-ing analgesics promote dependence? a review and evaluation.Clin Pharmacol Ther 68:457–467

Fennelly M, Galletly DC, Purdie GI (1991) Is caffeine withdrawalthe mechanism of postoperative headache? Anesth Analg72:449–453

Field AS, Laurienti PJ, Yen YF, Burdette JH, Moody DM (2003)Dietary caffeine consumption and withdrawal: confoundingvariables in quantitative cerebral perfusion studies? Radiology227:129–135

Fillmore MT (1994) Investigating the behavioral effects of caffeine:the contribution of drug-related expectancies. Pharmacopsy-choecologia 7:63–73

Fredholm BB, Bättig K, Holmen J, Nehlig A, Zvartau EE (1999)Actions of caffeine in the brain with special reference to factorsthat contribute to its widespread use. Pharmacol Rev 51:83–133

Galletly DC, Fennelly M, Whitwam JG (1989) Does caffeinewithdrawal contribute to postanaesthetic morbidity? Lancet1:1335

Garrett BE, Griffiths RR (1998) Physical dependence increases therelative reinforcing effects of caffeine versus placebo. Psycho-pharmacology (Berl) 139:195–202

Gibson CJ (1981) Caffeine withdrawal elevates urinary MHPGexcretion. N Engl J Med 304:363

Gilbert RJ (1986) Caffeine, the most popular stimulant. ChelseaHouse, New York

Gilbert RM (1984) Caffeine consumption. In: Spiller GA (ed) Themethylxanthine beverages and foods: chemistry, consumption,and health effects. Alan R. Liss, New York, pp 185–213

Goldstein A (1964) Wakefulness caused by caffeine. Arch ExpPathol Pharmakol 248:269–278

Goldstein A, Kaizer S (1969) Psychotropic effects of caffeine inman. III. A questionnaire survey of coffee drinking and itseffects in a group of housewives. Clin Pharmacol Ther 10:477–488

Goldstein A, Wallace ME (1997) Caffeine dependence in school-children? Exp Clin Psychopharmacol 5:388–392

Goldstein A, Kaizer S, Whitby O (1969) Psychotropic effects ofcaffeine in man. IV. Quantitative and qualitative differencesassociated with habituation to coffee. Clin Pharmacol Ther10:489–497

Greden JF (1974) Anxiety or caffeinism: a diagnostic dilemma. AmJ Psychiatry 131:1089–1092

Greden JF, Fontaine P, Lubetsky M, Chamberlin K (1978) Anxietyand depression associated with caffeinism among psychiatricinpatients. Am J Psychiatry 135:963–966

Greden JF, Victor BS, Fontaine P, Lubetsky M (1980) Caffeine-withdrawal headache: a clinical profile. Psychosomatics21:411–413 (see also pages 417 and 418)

Griffiths RR, Mumford GK (1996) Caffeine reinforcement, dis-crimination, tolerance, and physical dependence in laboratoryanimals and humans. In: Schuster CR, Kuhar MJ (eds)Pharmacological aspects of drug dependence: toward anintegrated neurobehavioral approach (Handbook of Experi-mental Pharmacology). Springer, Berlin Heidelberg New York,pp 315–341

Griffiths RR, Woodson PP (1988) Caffeine physical dependence: areview of human and laboratory animal studies. Psychophar-macology (Berl) 94:437–451

Griffiths RR, Bigelow GE, Liebson IA (1986) Human coffeedrinking: reinforcing and physical dependence producingeffects of caffeine. J Pharmacol Exp Ther 239:416–425

Griffiths RR, Evans SM, Heishman SJ, Preston KL, Sannerud CA,Wolf B, Woodson PP (1990) Low-dose caffeine physicaldependence in humans. J Pharmacol Exp Ther 255:1123–1132

Griffiths RR, Juliano LM, Chausmer AL (2003) Caffeine pharma-cology and clinical effects. In: Graham AW, Schultz TK, Mayo-Smith M, Ries RK, Wilford BB (eds) Principles of addictionmedicine, 3rd edn. American Society of Addiction Medicine,Chevy Chase, pp 193–224

Hale KL, Hughes JR, Oliveto AH, Higgins ST (1995) Caffeine self-administration and subjective effects in adolescents. Exp ClinPsychopharmacol 3:364–370

Hampl KF, Stotz G, Schneider MC (1994) Postoperative transienthemihypaesthesia and severe headache associated with caffeinewithdrawal. Anaesthesia 49:266–267

Hampl KF, Schneider MC, Ruttimann U, Ummenhofer W, Drewe J(1995) Perioperative administration of caffeine tablets forprevention of postoperative headaches. Can J Anaesth 42:789–792

Hindmarch I, Quinlan PT, Moore KL, Parkin C (1998) The effectsof black tea and other beverages on aspects of cognition andpsychomotor performance. Psychopharmacology (Berl)139:230–238

Höfer I, Bättig K (1994a) Cardiovascular, behavioral, and subjectiveeffects of caffeine under field conditions. Pharmacol BiochemBehav 48:899–908

Höfer I, Bättig K (1994b) Psychophysiological effects of switchingto caffeine tablets or decaffeinated coffee under field condi-tions. Pharmacopsychoecologia 7:169–177

Horst K, Buxton RE, Robinson WD (1934) The effect of thehabitual use of coffee or decaffeinated coffee upon bloodpressure and certain motor reactions of normal young men. JPharmacol Exp Ther 52:322–337

Hughes JR (1994) Caffeine withdrawal, dependence, and abuse. In:American psychiatric association: diagnostic and statisticalmanual of mental disorders, 4th edn. American PsychiatricAssociation, Washington, pp 129–134

27

Hughes JR, Oliveto AH (1997) A systematic survey of caffeineintake in Vermont. Exp Clin Psychopharmacol 5:393–398

Hughes JR, Higgins ST, Bickel WK, Hunt WK, Fenwick JW,Gulliver SB, Mireault GC (1991) Caffeine self-administration,withdrawal, and adverse effects among coffee drinkers. ArchGen Psychiatry 48:611–617

Hughes JR, Hunt WK, Higgins ST, Bickel WK, Fenwick JW,Pepper SL (1992) Effect of dose on the ability of caffeine toserve as a reinforcer in humans. Behav Pharmacol 3:211–218

Hughes JR, Oliveto AH, Bickel WK, Higgins ST, Badger GJ (1993)Caffeine self-administration and withdrawal: incidence, indivi-dual differences and interrelationships. Drug Alcohol Depend32:239–246

Hughes JR, Oliveto AH, Bickel WK, Higgins ST, Badger GJ (1995)The ability of low doses of caffeine to serve as reinforcers inhumans: A replication. Exp Clin Psychopharmacol 3:358–363

Hughes JR, Oliveto AH, Liguori A, Carpenter J, Howard T (1998)Endorsement of DSM-IV dependence criteria among caffeineusers. Drug Alcohol Depend 52:99–107

James JE (1997) Understanding caffeine. Sage, Thousand OaksJames JE (1998) Acute and chronic effects of caffeine on

performance, mood, headache, and sleep. Neuropsychobiology38:32–41

Jones HE, Herning RI, Cadet JL, Griffiths RR (2000) Caffeinewithdrawal increases cerebral blood flow velocity and altersquantitative electroencephalography (EEG) activity. Psycho-pharmacology (Berl) 147:371–377

Juliano LM, Brandon TH (2002) Effects of nicotine dose,instructional set, and outcome expectancies on the subjectiveeffects of smoking in the presence of a stressor. J AbnormPsychol 111:88–97

Kendler KS, Prescott CA (1999) Caffeine intake, tolerance, andwithdrawal in women: a population-based twin study. Am JPsychiatry 156:223–228

Kenemans JL, Wieleman JS, Zeegers M, Verbaten MN (1999)Caffeine and stroop interference. Pharmacol Biochem Behav63:589–598

Kingdon (1883) Effects of tea and coffee drinking. Lancet II:47–48Lader M, Cardwell C, Shine P, Scott N (1996) Caffeine withdrawal

symptoms and rate of metabolism. J Psychopharmacol 10:110–118

Lane JD (1994) Neuroendocrine responses to caffeine in the workenvironment. Psychosom Med 56:267–270

Lane JD (1997) Effects of brief caffeinated-beverage deprivation onmood, symptoms, and psychomotor performance. PharmacolBiochem Behav 58:203–208

Lane JD, Phillips-Bute BG (1998) Caffeine deprivation affectsvigilance performance and mood. Physiol Behav 65:171–175

Lieberman HR, Wurtman RJ, Emde GG, Coviella IL (1987) Theeffects of caffeine and aspirin on mood and performance. J ClinPsychopharmacol 7:315–320

Liguori A, Hughes JR (1997) Caffeine self-administration inhumans: 2. A within-subjects comparison of coffee and colavehicles. Exp Clin Psychopharmacol 5:295–303

Liguori A, Hughes JR, Grass JA (1997a) Absorption and subjectiveeffects of caffeine from coffee, cola and capsules. PharmacolBiochem Behav 58:721–726

Liguori A, Hughes JR, Oliveto AH (1997b) Caffeine self-administration in humans: 1. Efficacy of cola vehicle. ExpClin Psychopharmacol 5:286–294

Mackenzie TB, Popkin MK, Dziubinski J, Sheppard JR (1981)Effects of caffeine withdrawal on isoproterenol-stimulatedcyclic adenosine monophosphate. Clin Pharmacol Ther30:436–438

Marlatt GA, Rohsenow DJ (1980) Cognitive processes in alcoholuse: expectancy and the balanced placebo design. In: Mello NK(ed) Advances in substance abuse: behavioral and biologicalresearch. JAI, Greenwich, pp 159–199

Martin WR (1977) Drug addiction I: morphine, sedative/hypnoticand alcohol dependence. Springer, New York Berlin Heidelberg

Mathew RJ, Wilson WH (1985) Caffeine consumption, withdrawaland cerebral blood flow. Headache 25:305–309

McGowan JD, Altman RE, Kanto WP (1988) Neonatal withdrawalsymptoms after chronic maternal ingestion of caffeine. SouthMed J 81:1092–1094

Mitchell SH, de Wit H, Zacny JP (1995) Caffeine withdrawalsymptoms and self-administration following caffeine depriva-tion. Pharmacol Biochem Behav 51:941–945

Mumford GK, Benowitz NL, Evans SM, Kaminski BJ, Preston KL,Sannerud CA, Silverman K, Griffiths RR (1996) Absorptionrate of methylxanthines following capsules, cola and chocolate.Eur J Clin Pharmacol 51:319–325

Naismith DJ, Akinyanju PA, Szanto S, Yudkin J (1970) The effect,in volunteers, of coffee and decaffeinated coffee on bloodglucose, insulin, plasma lipids and some factors involved inblood clotting. Nutr Metab 12:144–151

Nehlig A (1999) Are we dependent upon coffee and caffeine? Areview on human and animal data. Neurosci Biobehav Rev23:563–576

Nikolajsen L, Larsen KM, Kierkegaard O (1994) Effect of previousfrequency of headache, duration of fasting and caffeineabstinence on perioperative headache. Br J Anaesth 72:295–297

Oberstar JV, Bernstein GA, Thuras PD (2002) Caffeine use anddependence in adolescents: one-year follow-up. J ChildAdolesc Psychopharmacol 12:127–135

Oliveto AH, Hughes JR, Higgins ST, Bickel WK, Pepper SL, SheaPJ, Fenwick JW (1992a) Forced-choice versus free-choiceprocedures: caffeine self-administration in humans. Psycho-pharmacology (Berl) 109:85–91

Oliveto AH, Hughes JR, Pepper SL, Bickel WK, Higgins ST(1992b) Low doses of caffeine can serve as reinforcers inhumans. In: Problems of drug dependence, 1990 NIDAresearch monograph no. 178. U.S. Government Printing Office,Washington, p 442

Patrick G, Reeves RR, Struve FA (1996) Does caffeine cessationincrease firing rates of diffuse paroxysmal slowing dysrhyth-mia? A serendipitous observation. Clin Electroencephalogr27:78–83

Phillips-Bute BG, Lane JD (1998) Caffeine withdrawal symptomsfollowing brief caffeine deprivation. Physiol Behav 63:35–39

Rainey JT (1985) Headache related to chronic caffeine addiction.Tex Dent J 102:29–30

Reeves RR, Struve FA, Patrick G, Bullen JA (1995) Topographicquantitative EEG measures of alpha and theta power changesduring caffeine withdrawal: preliminary findings from normalsubjects. Clin Electroencephalogr 26:154–162

Reeves RR, Struve FA, Patrick G (1997) Somatic dysfunctionincrease during caffeine withdrawal. J Am Osteopath Assoc97:454–456

Reeves RR, Struve FA, Patrick G (1999) The effects of caffeinewithdrawal on cognitive P300 auditory and visual evokedpotentials. Clin Electroencephalogr 30:24–27

Reeves RR, Struve FA, Patrick G (2002) Topographic quantitativeEEG response to acute caffeine withdrawal: a comprehensiveanalysis of multiple quantitative variables. Clin Electroence-phalogr 33:178–188

Richardson NJ, Rogers PJ, Elliman NA, O’Dell RJ (1995) Moodand performance effects of caffeine in relation to acute andchronic caffeine deprivation. Pharmacol Biochem Behav52:313–320

Rippere V (1984) Some varieties of food intolerance in psychiatricpatients: an overview. Nutr Health 3:125–136

Rizzo AA, Stamps LE, Fehr LA (1988) Effects of caffeinewithdrawal on motor performance and heart rate changes. IntJ Psychophysiol 6:9–14

Robelin M, Rogers PJ (1998) Mood and psychomotor performanceeffects of the first, but not of subsequent, cup-of-coffeeequivalent doses of caffeine consumed after overnight caffeineabstinence. Behav Pharmacol 9:611–618

Robertson D, Wade D, Workman R, Woosley RL, Oates JA (1981)Tolerance to the humoral and hemodynamic effects of caffeinein man. J Clin Invest 67:1111–1117

28

Rogers PJ, Richardson NJ, Elliman NA (1995) Overnight caffeineabstinence and negative reinforcement of preference forcaffeine-containing drinks. Psychopharmacology (Berl)120:457–462

Rogers PJ, Martin J, Smith C, Heatherley SV, Smit HJ (2003)Absence of reinforcing, mood and psychomotor performanceeffects of caffeine in habitual non-consumers of caffeine.Psychopharmacology (Berl) 167:54–62

Roller L (1981) Caffeinism: subjective quantitative aspect ofwithdrawal syndrome. Med J Aust 1:146

Rubin GJ, Smith AP (1999) Caffeine withdrawal and headaches.Nutr Neurosci 2:123–126

Schuh KJ, Griffiths RR (1997) Caffeine reinforcement: the role ofwithdrawal. Psychopharmacology (Berl) 130:320–326

Silverman K, Evans SM, Strain EC, Griffiths RR (1992) Withdrawalsyndrome after the double-blind cessation of caffeine con-sumption. N Engl J Med 327:1109–1114

Sjaastad O, Bakketeig LS (2004) Caffeine-withdrawal headache: theVaga study of headache epidemiology. Cephalalgia 24:241–249

Smit HJ, Rogers PJ (2000) Effects of low doses of caffeine oncognitive performance, mood and thirst in low and highercaffeine consumers. Psychopharmacology (Berl) 152:167–173

Smith AP (1996) Caffeine dependence: an alternative view. Nat Med2:494

Smith R (1987) Caffeine withdrawal headache. J Clin Pharm Ther12:53–57

Strain EC, Griffiths RR (1998) Characteristics of patients withchronic use of OTC analgesics containing caffeine. In:Problems of drug dependence, 1997 NIDA research monographno. 178. U.S. Government Printing Office, Washington, p 286

Strain EC, Mumford GK, Silverman K, Griffiths RR (1994)Caffeine dependence syndrome. Evidence from case historiesand experimental evaluations. J Am Med Assoc 272:1043–1048

Streufert S, Pogash R, Miller J, Gingrich D, Landis R, Lonardi L,Severs W, Roache JD (1995) Effects of caffeine deprivation oncomplex human functioning. Psychopharmacology (Berl)118:377–384

Stringer KA, Watson WA (1987) Caffeine withdrawal symptoms.Am J Emerg Med 5:469

Swerdlow NR, Eastvold A, Gerbranda T, Uyan KM, Hartman P,Doan Q, Auerbach P (2000) Effects of caffeine on sensorimotorgating of the startle reflex in normal control subjects: impact ofcaffeine intake and withdrawal. Psychopharmacology (Berl)151:368–378

Thomas DB (1988) Neonatal abstinence syndrome. Med J Aust148:598

Tinley EM, Yeomans MR, Durlach PJ (2003) Caffeine reinforcesflavour preference in caffeine-dependent, but not long-termwithdrawn, caffeine consumers. Psychopharmacology (Berl)166:416–423

Van Soeren MH, Graham TE (1998) Effect of caffeine onmetabolism, exercise endurance, and catecholamine responsesafter withdrawal. J Appl Physiol 85:1493–1501

Verhoeff FH, Millar JM (1990) Does caffeine contribute topostoperative morbidity? Lancet 336:632

Victor BS, Lubetsky M, Greden JF (1981) Somatic manifestationsof caffeinism. J Clin Psychiatry 42:185–188

Watson JM, Lunt MJ, Morris S, Weiss MJ, Hussey D, Kerr D (2000)Reversal of caffeine withdrawal by ingestion of a soft beverage.Pharmacol Biochem Behav 66:15–18

Weber JG, Ereth MH, Danielson DR (1993) Perioperative ingestionof caffeine and postoperative headache. Mayo Clin Proc68:842–845

Weddington WW, Brown BS, Haertzen CA, Cone EJ, Dax EM,Herning RI, Michaelson BS (1990) Changes in mood, craving,and sleep during short-term abstinence reported by malecocaine addicts. A controlled, residential study. Arch GenPsychiatry 47:861–868

Weil A, Rosen W (1993) From chocolate to morphine: everythingyou need to know about mind-altering drugs. HoughtonMifflin, Boston

Winstead DK (1976) Coffee consumption among psychiatricinpatients. Am J Psychiatry 113:1447–1450

World Health Organization (1992a) The ICD-10 classification ofmental and behavioural disorders: clinical descriptions anddiagnostic guidelines. World Health Organization, Geneva

World Health Organization (1992b) International statistical classi-fication of diseases and related health problems, 10th revision.World Health Organization, Geneva

World Health Organization (1993) The ICD-10 classification ofmental and behavioural disorders: diagnostic criteria forresearch. World Health Organization, Geneva

Yeomans MR, Spetch H, Rogers PJ (1998) Conditioned flavourpreference negatively reinforced by caffeine in humanvolunteers. Psychopharmacology (Berl) 137:401–409

Yeomans MR, Jackson A, Lee MD, Nesic J, Durlach PJ (2000a)Expression of flavour preferences conditioned by caffeine isdependent on caffeine deprivation state. Psychopharmacology(Berl) 150:208–215

Yeomans MR, Jackson A, Lee MD, Steer B, Tinley E, Durlach P,Rogers PJ (2000b) Acquisition and extinction of flavourpreferences conditioned by caffeine in humans. Appetite35:131–141

Yeomans MR, Ripley T, Lee MD, Durlach PJ (2001) No evidencefor latent learning of liking for flavours conditioned by caffeine.Psychopharmacology (Berl) 157:172–179

Yeomans MR, Pryke R, Durlach PJ (2002a) Effect of caffeine-deprivation on liking for a non-caffeinated drink. Appetite39:35–42

Yeomans MR, Ripley T, Davies LH, Rusted JM, Rogers PJ (2002b)Effects of caffeine on performance and mood depend on thelevel of caffeine abstinence. Psychopharmacology (Berl)164:241–249

29

caffeine withdrawal article

Documents