Behavioral and physiological responses to anabolic - ResearchGate

24 Pages·2003·0.33 MB·English

NeuroscienceandBiobehavioralReviews27(2003)413–436
www.elsevier.com/locate/neubiorev
Review
Behavioral and physiological responses to anabolic-androgenic steroids
Ann S. Clarka,*, Leslie P. Hendersonb,c
aDepartmentofPsychologicalandBrainSciences,DartmouthCollege,6207MooreHall,Hanover,NH03755,USA
bDepartmentofPhysiology,DartmouthMedicalSchool,328RemsenBuilding,Hanover,NH03755,USA
cDepartmentofBiochemistry,DartmouthMedicalSchool,328RemsenBuilding,Hanover,NH03755,USA
Received21February2003;revised15May2003;accepted28May2003
Abstract
Anabolic-androgenicsteroids(AAS)aresyntheticderivativesoftestosteroneoriginallydesignedfortherapeuticusestoprovideenhanced
anabolic potency with negligible androgenic effects. Although AAS continue to be used clinically today, the medical benefits of low
therapeuticdosesofAASstandinsharpcontrasttothepotentialhealthrisksassociatedwiththeexcessivedosesself-administerednotonly
byeliteathletesandbodybuilders,butbyagrowingnumberofrecreationalusers,includingadolescentboysandgirls.Thedeleteriouseffects
ofAASonperipheralorgansandtheincidenceofalteredbehaviorsinAASabusershavebeenwelldocumentedinanumberofexcellent
currentreviews forclinical populations.However, acomparable synthesisof nonclinicalstudieshasnotbeenmade. Ourpurposeinthis
reviewistosummarizetheliteratureforanimalmodelsoftheeffectsofsupraphysiologicaldosesofAAS(e.g.thosethatmimichumanabuse
regimes)onbehaviorsandontheneuralcircuitryforthesebehaviors.Inparticular,wehavefocusedonstudiesinrodentsthathaveexamined
howAASalteraggression,sexualbehaviors,anxiety,reward,learning,andlocomotionandhowAASaltertheexpressionandfunctionof
neurotransmittersystemsandothersignalingmoleculesthatunderliethesebehaviors.
q2003ElsevierLtd.Allrightsreserved.
Keywords:Anabolicsteroids;Testosterone;Abuse;Rat,mouse;Hamster;Reproduction;Aggression,anxiety;Reward;Learningandmemory;Locomotion;
Neurotransmitter;Neurotransmitterreceptors
Contents
1. Introduction............................................................................. 414
2. Experimentaldesign....................................................................... 415
3. BehavioralcorrelatesofAASadministration..................................................... 416
3.1. Aggression ......................................................................... 416
3.2. Sexualbehaviors..................................................................... 419
3.2.1. Malesexualbehavior............................................................ 419
3.2.2. Femalesexualbehavior .......................................................... 420
3.3. Anxiety............................................................................ 420
3.4. Reward............................................................................ 422
3.5. Learningandmemory................................................................. 422
3.6. Locomotion......................................................................... 423
4. PhysiologicalcorrelatesofAASadministration................................................... 423
4.1. Androgenreceptors................................................................... 424
4.2. g-AminobutyricacidtypeA(GABA )receptors ............................................. 425
A
4.2.1. AllostericmodulationofGABA receptors............................................ 425
A
4.2.2. ChangesinGABA receptorexpression.............................................. 426
A
4.3. 5-Hydroxytryptamine(5-HT)and5-HTreceptors............................................. 427
4.4. Dopamineanddopaminereceptors........................................................ 428
4.5. Opioidsandopioidreceptors ............................................................ 429
4.6. Othermoleculesinreward.............................................................. 430
* Correspondingauthor.Tel.:þ1-603-646-3036;fax:þ1-603-646-1419.
E-mailaddress:[email protected](A.S.Clark).
0149-7634/03/$-seefrontmatterq2003ElsevierLtd.Allrightsreserved.
doi:10.1016/S0149-7634(03)00064-2
414 A.S.Clark,L.P.Henderson/NeuroscienceandBiobehavioralReviews27(2003)413–436
4.7. Otherneuropeptidesandneurotrophins..................................................... 430
4.8. Stresshormones ..................................................................... 431
5. Conclusionsandfuturedirections............................................................. 431
References ................................................................................ 432
1. Introduction half-lifeandthustheefficacyofsyntheticAAS.Threemain
classes of AAS have been described. The first, used
Anabolic-androgenic steroids (AAS) are synthetic primarily as injectable compounds, are derived from
derivatives of testosterone originally designed to provide esterification of the 17b-hydroxyl group of testosterone
enhancedanabolic(tissue-building)potencywithnegligible and includes testosterone propionate and testosterone
androgenic(masculinizing)effects[79,93].Approximately, cypionate. Esterification retards degradation and prolongs
60 different AAS are available that vary in their chemical the duration of action after injection of the hormone by
structureand thusintheir metabolicfate andphysiological slowing its release into circulation [8,132]. Testosterone
effects [8,75,82,121,129,132,156]. All AAS are thought to esters can be hydrolyzed into free testosterone, reduced to
have some androgenic activity, and the androgen receptor 5a-dihydrotestosterone,anandrogenwithhigherbiological
bindingpropertiesofseveralofthesecompoundshavebeen activity at brain androgen receptors than testosterone [79,
characterized in brain tissue [124]. The AAS, like the 97,156],oraromatizedtoestrogens[79,156].Moleculesthat
endogenous androgens, are four-ringed structures with 19 havebeen5a-reducedcannotbemetabolizedintoestrogens
carbon atoms (Fig. 1). Modifications of this androstane but may be metabolized into other androgenic compounds
backbone have been introduced to prolong the metabolic such as3a-androstanediol.
Thesecondclassisalsocomposedofinjectableandrogen
esters called 19-nor-testosterone derivatives. These com-
pounds have, in conjunction with the addition of long side
chain moieties, a substitution of a hydrogen for the methyl
group at C19 [8,132] (Fig. 1). AAS in this class include
nandrolone decanoate. The substitution at C19 extends the
half-life of this class of AAS beyond that contributed by
esterification alone [132]. It should be noted that despite
retention of the C4–C5 double bond in nandrolone
decanoate, this compound has reduced androgenic activity
at the androgen receptor compared to dihydrotestosterone
[126,132,156]. Nandrolone decanoate, like the testosterone
esters in class I, can be aromatized to 17b-estradiol, albeit
only ,20% as efficiently as testosterone [126,156].
Aromatizable AAS in both class I and class II may thus
have additional and significant central nervous system
effects not only at the androgenreceptor, butalso vis-a`-vis
theactionsoftheirestrogenicmetabolitesatbrainestrogen
receptors [154].
ThethirdclassofAAScomprisesthosecompoundsthat
are alkylated at C17, including 17a-methyltestosterone,
oxymetholone,methandrostenolone,andstanozolol(Fig.1).
Because alkylation retards metabolism by the liver, this
group of AAS is orally active [8]. To the best of current
knowledge,noneofthe17a-alkylatedsteroidsisconverted
into dihydrotestosterone or 17b-estradiol although other
androgenicandestrogenicmetabolitesmaybeformed[156].
AAS were originally developed for the treatment of
hypogonadal dysfunction in men, initiation of delayed
puberty,andgrowthpromotion[8,154].Theycontinuetobe
used today for these treatments, as well as for therapy in
chronic conditions including HIV/AIDS, cancer, severe
Fig.1.Chemicalstructuresofrepresentativeexamplesofthethreemajor
classesoftheAAS. burns,anemia,hepaticandkidneyfailure,breastcancer,and
A.S.Clark,L.P.Henderson/NeuroscienceandBiobehavioralReviews27(2003)413–436 415
hereditary angioedema [8]. Originally synthesized to have administer doses oftestosterone propionateandotherAAS
maximal effects on protein synthesis and muscle growth above 3mg/kg/animal/day and define these doses as
withminimalandrogenicormasculinizingeffects,itisnow supraphysiological. We have limited our discussion of
clear from biochemical studies that there are no pure studies that administer lower doses to those that illustrate
anabolic steroids that are devoid of androgenic actions [8, thecontrastingeffectsoftheselowdosestothoseproduced
93,154]. Moreover, although originally developed for byAASgivenathighdoses.Inaddition,ourreviewfocuses
clinical use, AAS administration is now predominantly on those AAS that humans report self-administering (in
oneofabuse,andthemedicalbenefitsoflowdosesofAAS published studies) and does not address the effects of
stand in sharp contrast to the potential health risks androgens that are not self-administered by humans.
associated with the excessive doses self-administered by Second, is the fact that AAS may be administered to
elite athletes [48,49,92]. Assessments made a decade ago laboratory animals following a variety of treatment regi-
[161]indicatethat,atthattime,morethanonemillionadult mens and that results will vary significantly depending on
Americans had or were using AAS to increase physical which specific regimen is employed. For example, some
strength, endurance, athletic ability or muscle mass. Over studies simultaneously administer combinations of two or
the past few decades, the misuse of AAS by college, threeAAS,so-calledAAS‘cocktails’,tomimicthepattern
Olympic, and professional athletes has become widely of combining or ‘stacking’ of AAS that humans use [82].
recognized [80]. However, recent reports highlight the fact Although the administration of combinations of AAS
thatthemoreinsidiousgrowthintheabuseofthesedrugsis reflects human abuse paradigms, this method of exposure
not among elite athletes, but among adolescent boys and confoundsinterpretationofresultsandtheunderstandingof
girls[116].Presentestimationsarethat,4%ofhighschool the mechanism(s) because it is not known how AAS
students have used AAS [7,78,116,162], and the greatest interactionsaffectbehaviorandphysiologicalresponses.In
increaseinAASuseoverthepastdecadehasbeenreported addition, AAS are heterogeneous in their effects on many
inadolescent girls [7]. endpoints, as illustrated by the fact that class I and II AAS
ThepotentialbehavioraleffectsofAASabuseinhuman (as defined above) may be aromatized and act at the
populationshavereceivedprominentcoverageinanumber estrogen receptor, whereas class III AAS are believed to
of excellent recent reviews [6,29,48,56,82,93,162], as well have minimal estrogen receptor actions [156]. Conversely,
as in the popular press [53,80,115,142]. However, less while analyses of the actions of single AAS may provide
attention has been paid to the effects of AAS on neural bettermechanisticinsights,theydonotasaccuratelymirror
circuitsthatunderliethesebehavioraleffects,asdetermined human patterns ofself-administration.
instudiesusinganimal models.Our purposeinthisreview Third, the timeframe of AAS exposure varies dramati-
is to summarize the literature with respect to studies cally. In the studies described here, treatment times varied
employinganimalmodelsinconjunctionwithparadigmsof from 30min to 6 months. The specific timeframe for AAS
AAS exposure that mimic human abuse regimes (i.e. effects on behavior, minutes versus days, may provide
supraphysiological doses). In particular, we have focused valuable information regarding the underlying mechan-
on studies that have examined how AAS alter aggression, isms. For example, behavioral responses to acute AAS
reproductive behaviors, anxiety, learning and memory, administration would be consistent with allosteric modu-
reward and locomotion in rodents and how AAS alter the lation or post-translational actions of AAS at membrane
expression and function of neuronal signaling molecules delimited receptors, whereas behavioral responses that are
that underlie these behaviors. evident only after long-term exposure may implicate
changesingeneexpressionmediatedbyclassicalandrogen
receptororestrogenreceptorsignalingpathways.However,
2. Experimental design such differences also make comparison across studies that
employ widely divergent AAS exposures difficult. An
There are a number of experimental design consider- overall conclusion that can be drawn from the studies
ationsthatwillhaveanimpactontheconclusionsdrawnby presented in this review is that dramatically different
investigators assessing the behavioral and physiological results may be obtained with different experimental
effects of AAS. First, is the dose of AAS and whether it paradigms. Because of this, we believe that accurate
represents therapeutic androgen replacement levels or interpretation of the studies presented in this review can
mimics humanself-administrationthatresultsinsupraphy- bestbemadewithinthecontextoftheexperimentaldesign
siological androgen levels (abuse conditions). The present and methodologies of these studies, and these details are
review focuses on studies that administer supraphysiologi- provided herein.
cal doses of AAS. Physiological doses of testosterone Fourth, experiments may be performed on gonadally
propionate, as assessed in gonadectomized male rats intact or gonadectomized animals. Numerous studies have
administered hormone replacement to maintain reproduc- shownthatAAShavesignificanteffectsonthegonads,levels
tive and aggressive behaviors, are ,1mg/kg [32]. For of gonadal steroids, and gonadotropins [154], however, we
thepurposesofthisreview,wehavefocusedonstudiesthat haveconcentratedonAASeffectsoncentralnervoussystem
416 A.S.Clark,L.P.Henderson/NeuroscienceandBiobehavioralReviews27(2003)413–436
structures,andhavenotincludedadiscussionofthesestudies Morerecently,Breueretal.[21]examinedtheeffectsof
in this review. Analysis of sexual behaviors in gonadally opponent status (intact or gonadectomized) and environ-
intact animalsprovides an excellent behavioral window on ment (home cage, opponent cage, neutral cage) on AAS-
the impact of the AAS on the hypothalamic-pituitary- induced aggression. Testosterone propionate, nandrolone
gonadalaxis.Thus,studiesingonadallyintactanimalswill decanoate, or stanozolol was administered to Long-Evans
reflectboththenaturalstateandtheintricateinteractionsof rats (5mg/kg 5 times per week for 12 weeks). Aggressive
the exogenous AAS with the varying endogenous steroid behavior was increased in testosterone propionate-treated
milieu. However, the presence of circulating endogenous ratsrelativetocontrols,whereasthenandrolonedecanoate-
steroids can also confound results and make it difficult to treated and control groups exhibited similar levels of
interpret the critical neural substrates for AAS-modified aggression. Surprisingly, stanozolol-treated rats exhibited
behavior.Forthisreason,gonadectomizedanimalshavealso significantly lower levels of aggression than rats receiving
provedtobeusefulindelineatingthemechanismsunderlying not only testosterone propionate or nandrolone decanoate,
theeffectsofAASonbehavior. but also for those receiving vehicle. Rats receiving
Finally,significantdifferencesinAASactionsmayarise testosterone propionate continued to be responsive to the
not only from differences in hormonal state (e.g. intact or gonadal status of the opponent male, and thus were more
gonadectomized), but also with sex, age and species, and likelytoattackanintactversusagonadectomizedopponent,
eventhestrainofspeciesstudied.Althoughthereareonlya demonstrating that the testosterone propionate did not
few studies that have compared how sex, age or strain interfere with the ability to respond appropriately to these
modify the actions of AAS, studies abound that categori- social cues. Regarding the environmental manipulation,
cally demonstrate that brain function varies significantly whereas control rats displayed heightened aggression only
with sex, age, and genetic background [42,67,153]. More- in the home cage relative to the neutral cage, testosterone
over,studiesinanimalmodelsthatdirectlyassesshowsex, propionate-treated rats showed elevated aggression in both
ageandstrainmodulatetheimpactofAASonthebrainand the opponent’s cage and in the home cage. The authors
on behavior are particularly needed given reports from hypothesize that the administration of supraphysiological
humansubjectsthatindicatethatsomeoftheeffectsofAAS levels of testosterone propionate increased the propensity
induced prior to puberty may not be reversible upon foraggressioninmalerats,whilemaintainingtheabilityto
cessation of drug use [48,159] and that long-term con- respond appropriately to social cues. Another noteworthy
sequences from AAS abuse may be greater in women than finding from this study was that stanozolol suppressed the
inmen [43,48,66,68,138]. display of aggressive behavior. The paradoxical effects of
Given these experimental considerations, it should be stanozololhighlightthedisparityofactionofdifferentAAS
notedthatthemajorityofstudiesperformedtodatehasused and the need to understand the molecular mechanisms of
adultgonadallyintactmalerats.Althoughwediscussresults each compound inthe brain.
fromstudiesusingothersubjects,forclarityofpresentation, McGinnis et al. [106] tested whether a different type of
we have not overtly designated each of these studies as provocation, mild physical provocation using a tail pinch,
having been performed on gonadally intact male rats, and evokesanexaggeratedaggressiveresponseinratsreceiving
the reader should assume that studies were performed in AAS and whether physical provocation impairs the ability
these animals unless otherwise designated. of AAS-treated rats to discriminate between appropriate
socialandenvironmentalcues.Ratsweretreatedfollowing
thesamethreeAASconditionsasabove[21],5daysaweek
3. Behavioral correlates of AAS administration for12weeks,andtestedforintermaleaggressionbeginning
onweek12.Eachratunderwentabrief(1s)tailpinchand
3.1. Aggression thenwastestedunderhomecage,opponentcageandneutral
cageconditionswithanintactorgonadectomizedopponent
Androgens have long been recognized as modulators of male. Unlike controls, testosterone propionate-treated rats
aggression in male rats [5,141]. Many studies of AAS thatreceivedatailpinchexhibitedanincreaseinaggression
effectsonaggressionhavefocusedonintermaleaggression, inallsocialandenvironmentalconditions;dominatingtheir
a pattern of aggressive behavior that is dependent on the opponent whether the opponent was intact or gonadecto-
presence of androgens [30]. Intermale aggression can be mizedandregardlessofthetestingcondition(home,neutral
measured by systematically assessing the quality and or opponent’s cage). In contrast, controls that received a
quantity of aggressive acts displayed by a ‘resident’ male tail-pinchshowedmoreaggressionintheirhomecagethana
towards a strange or ‘intruder’ male. Lumia et al. [94] neutral cage, and more aggression toward gonadally intact
reported that long-term exposure (10 weeks) of gonadally opponents. Thus, the social and environmental cues that
intact Long-Evans rats totestosterone propionate (1mg/rat normallymodulateaggressivebehaviorwereovershadowed
3 times per week) increased the display of dominance bythechronicadministrationofhighlevelsoftestosterone
posturesandthreats,andreducedthenumberofsubmissive propionate.Inaddition,testosteronepropionate-treatedrats
postures, relative to controls. also showed increased aggression when their opponents
A.S.Clark,L.P.Henderson/NeuroscienceandBiobehavioralReviews27(2003)413–436 417
were administered a tail pinch. The authors speculate that the aggression scores for rats treated with 17a-methyltes-
testosteronepropionatesensitizestherattoitssurroundings tosterone were equivalent to the levels displayed by rats
and lowers the threshold to respond to provocation with treated with physiological doses of testosterone propionate
aggression. The findings of McGinnis et al. [106] also (400mg/rat) on most of the behavioral indices assessed. In
confirmedthosereportedinBreueretal.[21]demonstrating contrast, stanozolol failed to elicit aggressive behavior in
a lack of effect of nandrolone decanoate on aggressive gonadectomized rats. These findings illustrate that AAS-
behaviorandasuppressionofaggressioninratstreatedwith induced aggression is compound-specific in gonadecto-
stanozolol. mized, aswell as inintact, male rats.
In one of the few reports to assess behavior after BonsonandWinter[18]testedtheeffectsoftestosterone
withdrawal from AAS, McGinnis et al. [107] recently propionate (30mg/kg/day) on competitive aggression. In
examined whether AAS-enhanced aggression is reversible. thistask,pairsofFischerratsweretrainedtocompetefora
Intermale aggression was assessed in rats following with- foodrewardandapatternofdominanceisestablished.Once
drawal from AAS (testosterone propionate, nandrolone a stable dominance hierarchy was observed, testosterone
decanoate or stanozolol; 5mg/kg 5 days a week for 12 propionatewasadministeredtothenon-dominantratatthe
weeks). The experimental conditions (home cage, neutral conclusionofeachday’scompetitionsession.After14days
cage, opponent cage) and opponent status (intact or oftestosteronepropionate,thepreviouslynon-dominantrats
gonadectomized)werevariedontestsconductedbeginning increased their level of competition compared to non-
either 3 (acute condition; weeks 3–9) or 12 weeks (long- dominantratsreceivingthevehicle.Thatis,administration
termcondition;weeks12–18)afterwithdrawal.Generally, of testosterone propionate reversed the dominance hier-
ontheacutetests(weeks3–9),moreratsinthetestosterone archy.Inagreementwiththeseresults,MitchellandWilson
propionateconditionthaninthecontrolconditioncontinued [114] reported that administration of unesterified testoster-
to display threats and mounts. Aggressive behavior in rats one (100mg pellets) for 90 days, resulting in plasma
treated with nandrolone decanoate did not differ from testosteronelevelsof,10ng/mlsignificantlyimprovedthe
controls whereas stanozolol-treated rats tended to display successrateofratstestedonacompetitiontaskinwhichthe
slightly lower levels of aggression than controls. By 18 incentive was copulation with a sexually receptive female.
weeks withdrawal from AAS, aggressive behavior in the Lindqvist et al. [88] also examined competitive aggression
testosterone propionate- and stanozolol-treated rats had inratstreatedwithnandrolonedecanoate(15mg/kg)for14
returned to control levels. Thus, the effects of testosterone days. Beginning one week after the last nandrolone
propionateandstanozololonaggressivebehaviorappearto decanoate injection, the Wistar rats were acclimated to
be dependent upon the continued presence of the AAS. In newcagesandcagematesfor3days.Onday4,waterintake
addition, the withdrawal from testosterone propionate did was restricted and upon introduction of an active water-
notitselfinduceaggressivebehavior.Toourknowledgeno spout, the competitive aggressive behaviors displayed by
other studies have assessed aggressive behavior after long- AASandvehicle-treatedratswererecorded.Atsomepoint
term withdrawal from AAS. after the competition test, the rats then were placed in
In contrast to the aforementioned studies from McGin- individual cagesfor thedetermination ofbaseline levelsof
nis’sgroup,Longetal.[90]reportedthatSprague–Dawley water intake. The AAS-treated rats spent more time
rats receiving nandrolone decanoate (2mg/day/rat or drinkingduringthecompetitiontaskthancontrols,although
20mg/week/rat for 4 weeks) displayed heightened levels baseline drinking did not differ between the groups. These
ofaggressionrelativetocontrols.Thebasesfordiscrepancy findings are difficult to interpret because tests for competi-
between the findings of Long et al. [90] and more recent tive aggression took place 1 week after cessation of
reports [21,106] are not clear, but may be related to nandrolone decanoate treatment and the authors do not
differences in rat strain (Sprague–Dawley versus Long- provide information on the clearance rate for nandrolone
Evans) and/or specific testing conditions; Long et al. [90] decanoate.
provided 2weeks ofexperiencewith aggression beforethe The effects of long-term treatment with stanozolol [99]
AAS treatments began. Rats in McGinnis et al. [106] were and testosterone propionate [100] have also been tested in
not experienced in aggression before receiving AAS. The miceusinganisolation-inducedmodelofaggression.After
fact that the lack of effect of nandrolone decanoate on 3weeksofhousinginisolation,thesubjectwaspairedwith
intermale aggression was replicated [21,106] demonstrates an anosmic stranger gonadally intact male mouse in a
thatunderspecifictestingconditionsnandrolonedecanoate neutral cage and the display of aggressive behaviors was
does notinduce aggressionin Long-Evansmale rats. monitored. In the first set of studies [99], stanozolol was
In addition to studies in intact rats, the aggression- administered to Alderly Park mice at high (7mg/kg),
inducing properties of AAS have also been examined in medium(0.7mg/kg)orlow(0.07mg/kg)dosesonalternate
gonadectomized Long-Evans rats [32]. 17a-Methyltestos- daysforathree-weekperiod,andtestsforisolation-induced
terone(3mg/day)orstanozolol(400mg/day)wasadminis- aggressionwereconducted24hafterthelastinjection.Two
tered daily for 6 weeks to rats with tests for intermale cohortsofmiceweretested;aperi-pubertal(29-dayold)and
aggression conducted on weeks 3–6. On week 6, anadult(56-dayold)group.Stanozolol,atanydosetested,
418 A.S.Clark,L.P.Henderson/NeuroscienceandBiobehavioralReviews27(2003)413–436
hadnoeffectsonaggressivebehaviorineitheragegroup,in accompanying AAS treatment. Specifically, an arginine
agreement with the findings from intact [21,106] and vasopressin type 1A antagonist (OH-Phaa-D-Tyr(Me)-Phe-
gonadectomized[32]rats.Inalaterstudy,a rangeofdoses Gln-Asn-Arg-Pro-Arg-NH ) was administered into the
2
of testosterone propionate was administered to OF-1 mice anterior hypothalamus of hamsters pre-treated with the
weekly for 10 weeks (3.75–30mg/kg) [100] and isolation- AAS combination described above. AAS-treated hamsters
inducedaggressionwasmeasuredduringweeks8–10,24h receiving this antagonist exhibited a reduced number of
after the weekly testosterone propionate injection. Surpris- bites and attacks, but continued to initiate biting with the
ingly,fewbehavioraleffectsoftestosteronepropionatewere same latency as hamsters receiving the vehicle. These
observed, with no effects on attack or threat frequency. On results suggested a role for arginine vasopressin in the
week 10, a significant reduction in latency to attack was intensity and frequency, but not in the initiation, of
observed in subjects receiving 3.75, 7.5 and 30mg/kg aggressive behavior [62]. More recently, Grimes and
testosteronepropionate,butnotinsubjectsreceiving15mg/ Melloni [55] tested the effects of the selective serotonin
kgtestosteronepropionate.Theauthorsspeculatethatthelack reuptake inhibitor, fluoxetine, on the expression of AAS-
of responsiveness to androgens may reflect a reduced induced aggression. Fluoxetine (20mg/kg), administered
sensitivityoftheOF-1straintothehormone-dependenceof
1h before the aggression test, blocked AAS-induced
aggression. Although this hypothesis remains to be tested
aggression. Specifically, AAS-treated hamsters receiving
directly, it is known that different mouse strains vary
fluoxetinedisplayedfewerattacksandtooklongertoinitiate
significantly with respect to their sensitivity to endogenous
the first attack sequence relative to AAS-treated hamsters
gonadal steroids [137], strongly suggesting that genetic
receivingthevehicle.Thesefindingsareconsistentwiththe
backgroundwillalsoinfluencesensitivitytoAAS.
results of Bonson et al. [18,19] described earlier, and
Additional studies in mice have assessed aggressive
demonstrate that one mechanism by which AAS modulate
behaviorfollowingexposuretoAAScombinations[22,23].
aggression is via alterations in serotonin transmission.
Bronson [22] administered silastic capsules containing
Given that Melloni et al. [110] used a combination of
testosterone, testosterone cypionate, 17a-methyltestoster-
AAS that includes nandrolone, and that nandrolone
one and norethandrolone to intact male and female CF-1
decanoate failed to induce aggression in male rats [21,
miceatlowandhighdosesfor6months.Thelatencytoattack
106] it would be interesting to test whether nandrolone
andthelatencytoaccumulate10soffightingwererecorded.
decanoateortestosteronepropionatealoneelicitsaggressive
AAS (low or high dose) had no effect on aggression in the
behaviorinhamsters.Thispointagainraisestheminimally
malemice,perhapsbecauseofthehighincidence(9/10)of
explored issue of comparing the effects of AAS given
fighting in the control group. In contrast, female mice
individually versus in combination. In addition, a dose–
receivingtheAASatloworhigh(5-foldthelowdose)doses
response analysis to any AAS, individually or in combi-
exhibitedanincreasednumberofattacksandwerealsomore
nation, has yet to be conducted in hamsters. Finally, given
likely than control females to fight back when attacked.
the inhibitory effects of stanozolol on the expression of
Similar results were observed in a follow-up study that
aggressioninintactmalerats[21]itwouldbeinterestingto
focusedexclusivelyonfemalemice[23].Toourknowledge,
test whether stanozolol antagonizes AAS-induced aggres-
Bronson[22]andBronsonetal.[23]aretheonlystudiesto
sion inhamsters.
quantifyAASeffectsonaggressioninfemalesubjects.
In summary, the effects of AAS on aggression are sex-,
Melloni and colleagues [109,110] have conducted a
species- and compound-specific. The administration of
series of experiments examining the modulation of inter-
testosterone propionate at supraphysiological doses for a
male aggression in gonadally intact adolescent male
long period of time consistently enhanced aggression in
hamsters treated with a cocktail containing 2mg/kg
intact male rats [21,93,106]. Aggression in testosterone
testosterone cypionate, 2mg/kg nandrolone decanoate,
and1mg/kgboldenoneundecylenate(a19-nor-testosterone propionate-treated rats may be provoked more readily by
ester). Each hamster received the AAS combination physicalstimuli(tail-pinch),andratsreceivingtestosterone
beginning on postnatal day 27 for 14–30 consecutive propionate exhibited increases in aggression in social and
days, depending on the specific study. Aggression was environmental contexts that do not provoke aggression in
measured using the resident-intruder paradigm test on the controls [21,106]. Other AAS (stanozolol) either failed to
day after the final AAS injection. Hamsters receiving 14 stimulateaggression[32]oractuallyinhibitedthedisplayof
daysofAASdisplayedmorebitesandattacksandareduced aggression[21,106].Estrogens,aswellastestosterone,can
latencytobitetheintruderrelativetocontrols[109,110].In stimulateaggressivebehaviorsinrats[30]andaggressionis
subsequent studies by this lab, a similar increase in the alteredinmicewithtargeteddeletionsinthegeneencoding
numbersofbitesandattacks,andreducedbitelatency,was the estrogen receptor alpha [117]. It is intriguing to
observedinhamsterstreatedwiththeAAScombinationfor speculate that the differential effects of testosterone
30 days [41,55,62]. propionate, 17a-methyltestosterone, nandrolone decanoate
Harrison et al. [62] tested the hypothesis that arginine and stanozolol on aggression may reflect differences in the
vasopressinsystemmaymediatetheincreaseinaggression abilitiesofthesecompoundstoactatandrogenandestrogen
A.S.Clark,L.P.Henderson/NeuroscienceandBiobehavioralReviews27(2003)413–436 419
receptorsandresultingdifferencesinthebalanceofestrogen at either low or high doses on the display of male sexual
and androgen receptor-mediatedsignaling [106]. behaviorinadultCF-1 mice.
The studies done by Bronson and colleagues [22,23] ItisnoteworthythatinthestudybyClarketal.[37]the
suggest that AAS induce striking effects on aggression in AAS treatments that eliminated the expression of behavior
female mice, however, surprisinglylittle research has been in intact male rats (17a-methyltestosterone, stanozolol,
conducted on AAS effects on aggression in other female oxymetholone) suppressed serum testosterone levels,
animal models. The neural systems underlying the AAS whereas testosterone levels were not suppressed in rats
induction of aggression appear to overlap with the brain receiving high doses of the AAS that did not affect sexual
circuits underlying the regulation of aggression by behavior (testosterone cypionate, nandrolone decanoate,
endogenous androgens, that is neural systems utilizing methandrostenolone). Thus, the possibility cannot be
arginine vasopressin, serotonin, and GABA [20,47,57] and excluded that AAS effects on sexual behavior in intact
AASeffectsontheseneuralsystemsarediscussedindetail male rats reflected solely the suppression of endogenous
below. testosteronesecretiondistinctfromanydirectactionofAAS
compoundsinthebrainareasmodulatingmalereproductive
3.2. Sexual behaviors behavior.TodirectlyassesstheabilityandpotencyofAAS
toelicitmalesexualbehaviorintheabsenceofendogenous
3.2.1. Male sexual behavior testosterone, Clark and colleagues [34,36] exposed gonad-
AnumberofstudieshaveinvestigatedtheeffectsofAAS ectomizedsubjectstoAAS.AnalysesoftheabilityofAAS
on the sexual behavior of intact male rodents. Clark et al. to maintain sexual behavior following gonadectomy is a
[37]assessedtheeffectsof12weeksoftreatmentwiththree well-establishedparadigmforassessingtheandrogenicand
doses of each of six individual AAS on the expression of estrogenic potency of compounds [69]. AAS treatments
male sexual behavior in Long-Evans rats. The AAS and werebegunonthedayofgonadectomyandcontinueddaily
doses administered were as follows: 17a-methyltestoster- for six weeks with tests for sexual behavior conducted on
one (0.075, 0.75 and 7.5mg/kg); methandrostenolone weeks1–6.ThespecificAASanddosesadministeredwere
(0.0375, 0.375, 3.75mg/kg); nandrolone decanoate (0.056, asdescribedabove forintactmalerats [37].AASwerenot
0.56, 5.6mg/kg); stanozolol (0.05, 0.5, 5.0mg/kg), oxy- equipotent in maintaining male sexual behavior patterns in
metholone (0.125, 1.25, 12mg/kg); and testosterone gonadectomized male rats. Specifically, on week 6, 17a-
cypionate (0.075, 0.75, 7.5mg/kg). Twelve weeks of methyltestosterone,stanozolol,and oxymetholonefailedto
administrationofthehighdosesof17a-methyltestosterone, maintain sexual behavior in gonadectomized rats at any of
stanozolol,oroxymetholoneeliminatedthedisplayofmale thedosestested.ItisnoteworthythatthesearetheAASthat
sexual behavior; suppressing the expression of mounts, interfered with sexual behavior in intact male rats. In
intromissions and ejaculations. In contrast, methandroste- contrast, gonadectomized rats receiving the high dose of
nolone, nandrolone decanoate, and testosterone cypionate methandrostenolone, nandrolone decanoate or testosterone
had minimal effects on the display of sexual behaviors at cypionate continued to display male sexual behavior
any dose tested. Thus, in intact male rats, the six AAS throughout the 6-week testing period suggesting that these
examined in these studies evoked a range of behavioral AAS can substitute for endogenous testosterone and
responsesthatvariedasafunctionofthespecificcompound maintain the full expression ofmale sexualbehaviors.
anddoseadministered.Inagreementwiththelackofeffects In summary, AAS have heterogeneous effects on the
oflowdosesofAASonsexualbehaviorinintactmalerats, sexual behavior of male rats. AAS effects on sexual
Lumiaetal.[94]reportedthatlong-termexposureofLong- behavior in intact male subjects are largely absent except
Evans rats to testosterone propionate (1mg/rat 3 times per underperiodsofextendedexposureandhighdoses[37],and
week for 10 weeks) had no effects on the display of male low doses of AAS are generally without effect. Studies in
sexualbehaviors.Inaddition,Feinbergetal.[44]evaluated gonadectomized male rats show that AAS, which at high
the sexual behavior of Long-Evans rats administered doses disruptsexualbehavior inintactsubjects, alsofailto
testosterone propionate for 16 weeks (1mg/rat 3 times per maintain sexual behavior in gonadectomized subjects
week) beginning pre-pubertally (postnatal day 25) or peri- suggesting that these specific AAS (17a-methyltestoster-
pubertally(postnatalday40–45).Peri-pubertalexposureto one, stanozolol, and oxymetholone) fail to stimulate
testosteronepropionateincreasedtheproportionofratsthat androgen receptors in the brain to the level necessary to
displayedejaculation.Feinbergetal.[44]alsoexaminedthe support the display of male sexual behaviors. It is
sexualbehaviorofratsthatreceivedtestosteronepropionate noteworthy that no studies have been conducted in male
for3weeksandweretestedat16weeks(i.e.followinga13- ratsassessingsexualbehaviorsinresponsetocombinations
week withdrawal period) and determined that a smaller ofAAS.GiventhathumanmaleusersofAASoftenreport
proportionofratsinthetestosteronepropionate-withdrawal fertility complications such as decreased production of
group exhibited ejaculation compared with vehicle-treated testosteroneandsperm[4],itwouldbeworthwhiletoassess
rats.Finally,Bronson[22]observednosignificanteffectsof male reproductive function in animal subjects receiving
6monthsofexposuretoacombinationofAASadministered combinations oflow doses ofAAS.
420 A.S.Clark,L.P.Henderson/NeuroscienceandBiobehavioralReviews27(2003)413–436
3.2.2. Female sexualbehavior benzoate (2.0mg/day on days 1–19) to induce receptivity
The effects ofAAS onestrous cyclicity inboth rats and andadministeredeachofthesixAASmentionedabove(ata
micehavebeeninvestigated.Clarkandcolleagues[12a,38] dose of 7.5mg/kg) on days 7–19. Tests for sexual
administered six individual AAS (17a-methyltestosterone, receptivity were conducted on day 20 in the absence of
methandrostenolone, nandrolone decanoate, stanozolol, progesterone replacement. Two weeks of administration of
oxymetholone, testosterone cypionate) at low (0.037– 17a-methyltestosterone, methandrostenolone, nandrolone
0.075mg/kg), medium (0.375–0.75mg/kg) or high decanoate, or stanozolol inhibited sexual receptivity in
(3.75–7.5mg/kg) doses to intact Long-Evans female rats gonadectomized, estrogen-primed subjects. In contrast,
for 2 weeks. The doses of the individual AAS were as estrogen-induced receptivity was not inhibited in gonad-
described above in the studies of male rats [37]. Vaginal ectomized rats receiving either oxymetholone or testoster-
cytology and sexual behavior were monitored daily. Rats one cypionate. The inhibition of sexual receptivity in
receivingthehighdosesofallAASdisplayedsignificantly gonadectomized, hormone-primed female rats by AAS has
reduced periods of estrous cytology during the treatment beenshowntobedependentontheandrogenreceptor.The
periodversuscontrols.Somewhatsurprisingwasthefinding androgen receptor-specific antagonist, flutamide, comple-
thatAASaffectedsexualreceptivitydifferentlythanvaginal tely reversed the inhibitory effects of 17a-methyltestoster-
cytology,suggestingadifferentialsensitivitytoAASofthe one, stanozolol and nandrolone decanoate indicating that
neural substrates underlying the induction of sexual AAS actions at the androgen receptor are critical for the
receptivity in ovariectomized rats and those regulating suppressionofreceptivityingonadectomizedsubjects[13].
neuroendocrine function inthe intact rat. Therespectiverolesofsignalingmediatedbytheandrogen
Whereas the high doses of 17a-methyltestosterone, receptor and other potential targets of the AAS on the
methandrostenolone, and stanozolol all suppressed the disruptionofvaginalcytologyandonothersexualbehaviors
display of sexual receptivity, rats receiving the high doses (e.g., proceptive behaviors, estrous termination) have not
of nandrolone decanoate or oxymetholone continued to been determined.
display regular cycles of sexually receptive behavior Insummary,a short- (2week) orlong- (9months) term
withoutinterruptionduringthetreatmentperiod.Incontrast, administrationofAASdisruptstherodentestrouscycle.In
ratstreatedwithtestosteronecypionateshowedasignificant rats, there is evidence that sexual behavior, as well as
increase in the number of days of sexual receptivity peripheral cytology, is suppressed by some AAS. In mice,
displayed during the treatment period relative to controls. onlyvaginalcytologyhasbeenmonitored.Inaddition,AAS
The disruptive effects of 2 weeks of exposure to AAS on thatsuppressedthedisplayofsexualreceptivityinducedby
female sexual behavior were also found to be reversible in primingwithestrogeningonadectomizedratsdidsoviaan
most cases; regular cycles of vaginal cytology and sexual androgen receptor-dependent process. Collectively, these
receptivity resumed during the 2-week follow-up (with- results suggest that AAS act in the brain to interfere with
drawal) period from stanozolol, oxymetholone, 17a- events necessary for the estrogen-dependent induction of
methyltestosterone and methandrostenolone [12a,38]. In female sexual behavior and the regulation of the neuro-
contrast, rats receiving nandrolone decanoate or testoster- endocrine events requiredfor reproductive cyclicity.
one cypionate displayed changes in vaginal cytology and
sexual behavior that persisted through the 2-week with- 3.3.Anxiety
drawal period. These results highlight the heterogeneity of
AAS effects on both physiology and behavior and are one In contrast to the analysis of AAS effects on aggression
demonstration that the AAS that can be aromatized into and sexual behaviors, less is known about AAS effects on
estrogens (e.g. testosterone cypionate) may have distinct anxiety.AASeffectsonanxietywerefirstreportedbyBitran
physiologicalandbehavioraleffectsfromAASthatarenot et al. [10] who assessed the effects of a high dose of
subject toaromatization (e.g. stanozolol). testosterone propionate (silastic implants) on the perform-
In addition to studies in gonadally intact female rats, anceofLong-Evansratsontheelevatedplusmaze.Animals
combinations of four AAS (low and high doses) were weretestedateither6or14daysafterimplantsweremade.
administeredviasilasticcapsulestoCF-1miceforeither9 Serum levels of testosterone were elevated 7 to 10-fold in
weeks [23] or 6 months [22]. Not surprisingly, none of rats receiving implants relative to intact controls. Rats
AAS-treated mice exhibited vaginal estrous cycles at the having6daysoftestosteronepropionateexposuredisplayed
time ofassessment (9weeksor6months after initiationof an increase in the exploration of the open arms of the
AAS treatments). Sexual behaviors were not assessed in elevated plus maze relative to controls, a sign of reduced
these experiments, and as noted above [12a,39], sexual anxiety. However, the behavior of rats exposed to
behavior and vaginal cytology may be dissociated under testosterone propionate for 14 days did not differ from
sometreatments. controls. The differences in open arm exploration for rats
AAS effects on the induction of sexual receptivity have treated with testosterone propionate for 6 days were
also been assessed in gonadectomized Long-Evans female exhibited in the absence of any changes in open-field
rats[12].Gonadectomizedratswereadministeredestradiol activity, suggesting a specific effect of the testosterone
A.S.Clark,L.P.Henderson/NeuroscienceandBiobehavioralReviews27(2003)413–436 421
propionate treatment on anxiolytic behavior and not a thebehavioralandphysiologicaleffectsofthesesteroidscan
generalized effect on activitylevels. be attributedtochanges inGABA receptor.
A
Bitran et al. [10] suggest that the anxiolytic effects of FurtherevidencethatAASmayhaveanxiolyticeffectsis
AAS arise from metabolism of AAS to neurosteroid providedbythestudyofBingetal.[9]inwhichWistarrats
metabolites that induce allosteric modulation of the g- received a single injection of testosterone (unesterified,
aminobutyric acid type A (GABA receptor) and that the 5mg/kg) 24-h before testing on Vogel’s conflict test. On
A
bimodaleffectsofAASonanxietymeasuresmayreflectthe this test, water-deprived rats were trained to lick from a
development of tolerance at these receptors. The results spout inanoperant chamber.Followinga24-hdeprivation
suggesting development of tolerance to the anxiolytic period,theratswerereturnedtothechamberandallowedto
effects of AAS are intriguing given that AAS effects on drink without punishment. Thereafter each lick was
aggressionorsexualbehaviorsdonotappeartobetransient. accompaniedbyanelectricshockandthenumberofshocks
Perhaps this reflects the heterogeneity in AAS actions on received was recorded. Anti-anxiolytic agents increase
different neuronal signaling systems (e.g. the GABAergic punished drinking relative to controls [27], and Bing et al.
versus the serotonergic or androgen receptor-mediated [9] observed that testosterone-treated rats accepted signifi-
signaling). Although this is an attractive hypothesis, no cantlymoreshocksthancontrolrats,consistentwithananti-
studies have been performed to directly test the hypothesis anxiety action of testosterone.
that exposure to testosterone propionate results in changes It is noteworthy that Bing et al. [9] observe AAS-
in GABA receptor expression with a time course that dependent changes in anxiety within 24h after a single
A
parallels both the change in anxiolytic behavior and injection of testosterone. First, it is unlikely that anything
GABA receptor function reported by these authors. beyond physiological levels of testosterone would be
A
Moreover, no studies to date have examined the dose– present at 24h after a single injection, given the rapid
response relationship of testosterone propionate to the clearanceofunesterifiedtestosterone[154].However,ifthe
production of anxiolytic behaviors nor correlated the anxiolytic effects were due to this testosterone treatment,
expression of these behaviors with levels of neurosteroid then the narrow time frame, although not categorically
metabolitesinthebrain.Recentresultsdemonstratethatthe excluding androgen receptor-dependent changes in gene
AAS themselves (in the absence of metabolism to expression, would be more consistent with a nongenomic
neurosteroids) can allosterically modulate the GABA mechanism of AAS action. Evidence that androgens do
A
receptor (Fig. 2) [74,102,103,160]. Therefore, the findings indeedelicitrapidanxiolyticeffectshasbeenprovidedbya
ofBitranetal.[10]needtobeextendedtoestablishwhether recent study [2] on the effects of naturally occurring
the effects of AAS on anxiety behavior as measured in the androgens in adult, C57BL/6J£AKR/J hybrid male mice.
elevated plus maze can be attributed to modulation of In this study, the authors found that a single dose of either
GABA receptor function. Specifically, whether levels of testosterone (500mg: an amount given to reflect the
A
either neurosteroid metabolites or the AAS themselves pulsatile, elevated release of testosterone that is observed
reach concentrationsinthebrainthatwillleadtoallosteric during sexual encounters) or either of the testosterone-
modulation of these receptors has not been tested. derived neurosteroids, androsterone or 3a-androstanediol
Furthermore, it needs to be established that both the initial (100mg), reduced anxiety within 30min. The anxiolytic
enhancementandthesubsequenthypothesized tolerance to effects of testosterone were blocked by co-injection of
Fig.2.AllostericmodulationofneurotransmitterreceptorfunctionbytheAAS.(A)SchematicrepresentationoftheGABA receptor.TheGABA receptoris
A A
apentamericproteinwithaproposedstoichiometryof2a,2bandagsubunit.Multiplea(1–6),b(1–3)andg(1–3)subunitgenes,aswellassplicevariants
(notshown)havebeenidentified.Lessabundantsubunits(d,1 andQ)maysubstituteforthegsubunitinnativereceptors.Thecartoonillustratesthetwo
bindingsitesforGABAattheinterfaceoftheaandbsubunitsandaputativebindingsitefortheAAShypothesizedtoresideinthetransmembranedomain.
BindingoftheAASwiththissitemayinduceaconformationalchangeinthereceptor(!)thatalterstransitionsofthereceptorbetweendifferentkineticstates
wheneitheroneortwooftheGABAsitesarebound,thusproducingtheallostericmodulationoftheGABA-elicitedchloridecurrent.(B)Representative
currentsillustratingallostericmodulationofGABA receptorfunctionbytheAAS,17a-methyltestosterone(17a-MeT).Currentswereelicitedbyultrafast
A
perfusionof1mMGABA(EC ;2mspulseapplication)(seeRef.[160]formethods)torecombinantabg GABA receptorsexpressedinHEK293
95 2 3 2Long A
cells(courtesyofPaulYang;DartmouthMedicalSchool,Hanover,NH).Theeffectsof17a-MeT(increaseinpeakamplitudeandprolongationofcurrent
decay)areimmediateandreversible.Comparablepositivemodulationisobservedforsynapticcurrentselicitedfrombrainregionsexpressingpredominantly
abg GABA receptors[74].
2 3 2Long A
422 A.S.Clark,L.P.Henderson/NeuroscienceandBiobehavioralReviews27(2003)413–436
theGABA receptorantagonists,bicucullineorpicrotoxin, studiesindicatethatAASalterneuralsystemsimportantin
A
suggestingthattheabilityoftestosteronetoreduceanxiety reward.
arose from nongenomic allosteric modulation of this Theconditionedplacepreferencetaskisusedextensively
neurotransmitter receptor, although secondary effects on to evaluate the rewarding properties of drugs. Several
locomotorbehaviorobservedwithbicucullineorpicrotoxin studieshavereportedthattestosterone(,3mg/kg)doesnot
complicate interpretation of experiments with these antag- induceaconditionedplacepreferenceinmalerats[4,26,51,
onists.IncontrasttotheanxiolyticeffectsofAASreported 125c.f.,3].Therearenopublishedstudiestestingtheeffects
above, Minkin et al. [113] observed that Long-Evans rats ofotherAAS ordoses on the conditionedplace preference
(gonadally intact or gonadectomized) treated with nandro- task, nor has the AAS-induction of conditioned place
lonedecanoate(10or50mg/weekfor8weeks)spentmore preference been tested infemale subjects.
time than controls in the margins of the open-field (i.e. Giventhefrequentreferencesinthemediaandscientific
thigmotaxis),suggestinganincreased levelofanxiety.The literature to the abuse of AAS [24], surprisingly little
opposingresultsofMinkinetal.[113]versusBingetal.[9] evidence has been forthcoming demonstratingthat animals
and Aikey et al. [2] may reflect differences in AAS will self-administer AAS. Recently, Wood and colleagues
administered,dose andtime course. [73,158], using a food-induced drinking model that makes
In summary, the number of studies that have tested the testosteroneavailableinanoralsolution,demonstratedthat
effectsofAASonanxietybehaviorislimited.Thefindings gonadally intact adult male Siberian hamsters self-admin-
of an anxiety-reducing action of testosterone reported by ister testosterone up to 4mg/ml. Wood [158] reports
Bitran etal.[10]andBingetal.[9]requirefollow-upwith considerableindividualvariabilityinthepatternandamount
additional AAS, aswell as time course and dose–response of AAS self-administration among male hamster subjects,
characterizations. As with aggressive behaviors, it would and no evidence of a dose–response curve. Although the
also be valuable to test the effects of AAS on anxiety findings of Wood [158] are the clearest demonstration to
behaviors in female subjects. Given the actions of AAS at date that AAS are reinforcing, additional studies assessing
theGABA receptorasdescribedinSection4thatfollows the self-administration ofother AAS are needed.
A
and the established role of GABA in the expression of In summary, the recent findings of AAS self-adminis-
trationinhamstersprovideanexcellentmodelforthefuture
anxiolytic behaviors, the analysis of AAS effects on the
analyses of the abuse potential of AAS, and will allow for
GABAergic system and anxiety behaviors provides an
testing of different compounds, doses and time courses
opportunity for further experimental study. In this regard,
[158]. In addition, experiments testing AAS effects on the
the experiments reported by Aikey et al. [2] are quite
rewarding properties of other drugs (cocaine, heroin) may
intriguing, but reflect endogenous changes in testosterone
reveal interactions between AAS and commonly abused
and need to berepeated with AAS givenat high doses.
drugs. Finally, tests of the ability of supraphysiological
doses of testosterone propionate and other AAS to elicit a
3.4. Reward
conditionedplacepreferenceinmaleorfemalesubjectswill
extendourcharacterizationoftheabusepotentialofAASin
The potential reinforcing effects of AAS have been animal models.
investigatedusingseveralparadigms.Clarketal.[35]tested
whether AAS affect intracranial self-stimulation in Long- 3.5.Learning andmemory
Evans rats. Rats received daily injections of methandroste-
nolone(,3mg/kg)1hbeforedaily testingfor intracranial Aswithreward,fewstudieshaveexaminedtheeffectsof
self-stimulation for 14 consecutive days and showed no AAS on learning and memory. Two studies have assessed
changes in brain stimulation reward [35]. In a second spatial memory in male rats treated with AAS. Clark et al.
experiment, the effects of D-amphetamine (0.5mg/kg [33] administered an AAS combination (2mg/kg testoster-
15min prior to the testing session) on brain stimulation onecypionate,2mg/kgnandrolonedecanoate,and1mg/kg
rewardweredeterminedbeforeandaftertheratsreceiveda boldenone undecylenate), methandrostenolone (0.375mg/
combinationofAAS(2mg/kgtestosteronecypionate,2mg/ kg), or vehicle to Long-Evans rats for 10 weeks and tested
kg nandrolone decanoate, and 1mg/kg boldenone undecy- ratsinthewatermazetaskwiththeplatformhidden(place
lenate)dailyfor15weeks[35].Althoughbrainstimulation training)andremoved(probetrials).AAShadnoeffectson
reward remained stable in rats treated with this AAS performance on the water maze task. Rats receiving the
combination for 15 weeks, AAS-treated rats exhibited a combinationAASandmethandrostenolonewereabletofind
significantly larger response to amphetamine than vehicle- theplatformasquickly,butnomorequickly,thancontrols
treated rats [35]. These findings suggested that although andspentcomparabletimeswimminginthemazequadrant
AAS did not have direct effects on brain reward, AAS that previously housed the platform on the probe trials.
potentiatedtherewardingeffectsofamphetamine.Noother Smith et al. [135] tested AAS effects on a spatial working
studies have been conducted to determine whether AAS memory task. Long-Evans rats received 17a-methyltestos-
affect the rewarding properties of other drugs, although terone (7.5mg/kg), methandrostenolone (3.75mg/kg) or