Trait- and state-dependent cortical inhibitory deficits in bipolar disorder

Depósi o de In es igación de la Uni e sidad de Se illa h ps://idus.us.es/ This is he pee e iewed e sion o he ollowing a icle: Ruiz-Veguilla, M., Ma ín- Rod íguez, J. F., Paloma , F. J., Po cacchia, P., Ál a ez de Toledo, P., Pe ona-Ga celán, S., Rod íguez-Tes al, J. F., Hue as-Fe nández, I., & Mi , P. (2016). T ai - and s a e- dependen co ical inhibi o y de ici s in bipola diso de . Bipola Diso de s, 18(3), 261– 271, which has been published in inal o m a h ps://doi.o g/10.1111/bdi.12382 . This a icle may be used o non-comme cial pu poses in acco dance wi h Wiley Te ms and Condi ions o Use o Sel -A chi ed Ve sions. This a icle may no be enhanced, en iched o o he wise ans o med in o a de i a i e wo k, wi hou exp ess pe mission om Wiley o by s a u o y igh s unde applicable legisla ion. Copy igh no ices mus no be emo ed, obscu ed o modi ied. The a icle mus be linked o Wiley’s e sion o eco d on Wiley Online Lib a y and any embedding, aming o o he wise making a ailable he a icle o pages he eo by hi d pa ies om pla o ms, se ices and websi es o he han Wiley Online Lib a y mus be p ohibi ed. O iginal A icle T ai - and s a e-dependen co ical inhibi o y de ici s in bipola diso de Ruiz-Veguilla M, Ma  ın-Rod  ıguez JF, Paloma FJ, Po cacchia P,  Al a ez de Toledo P, Pe ona-Ga cel an S, Rod  ıguez-Tes al JF, Hue as-Fe n andez I, Mi P. T ai - and s a e-dependen co ical inhibi o y de ici s in bipola diso de . Bipola Diso d 2016: 00: 000–000. ©2016 John Wiley & Sons A/S. Published by John Wiley & Sons L d. Objec i es: Eu hymic pa ien s wi h bipola diso de (BD) ha e de ici s in co ical inhibi ion. Howe e , whe he co ical inhibi o y de ici s a e ai - o s a e-dependen impai men s is no ye known and hei ela ionship wi h psychia ic symp oms is no ye unde s ood. In he p esen s udy, we examined ai - and s a e-dependen co ical inhibi o y de ici s and e alua ed he po en ial clinical signi icance o hese de ici s. Me hods: Nine een pa ien s wi h bipola I diso de we e e alua ed using he pai ed-pulse ansc anial s imula ion p o ocol, which assessed co ical inhibi ion du ing an acu e manic episode. Co ical inhibi ion measu es we e compa ed wi h hose ob ained in 28 demog aphically ma ched heal hy con ols. A ollow-up assessmen was pe o med in 15 o hese pa ien s h ee mon hs la e , when he e was emission om hei mood and psycho ic symp oms. The associa ion be ween co ical inhibi o y measu es and se e i y o psychia ic symp oms was also s udied. Resul s: Du ing mania, pa ien s showed dec eased sho -in e al in aco ical and anscallosal inhibi ion, as well as a no mal co ical silen pe iod and long-in e al co ical inhibi ion. These indings we e he same du ing eu hymia. Symp oms associa ed wi h mo o hype ac i i y we e co ela ed nega i ely wi h he deg ee o co ical inhibi ion. These co ela ions we e no signi ican when a Bon e oni co ec ion was applied. Conclusions: The p esen longi udinal s udy showed co ical inhibi o y de ici s in pa ien s wi h BD, and suppo s he hypo hesis ha co ical inhibi o y de ici s in BD a e ai dependen . Fu he esea ch is necessa y o con i m he clinical signi icance o hese de ici s. Miguel Ruiz-Veguilla a,† ,Juan F ancisco Ma  ın-Rod  ıguez b, † , F ancisco J Paloma b ,Paolo Po cacchia b ,Paloma  Al a ez de Toledo b ,Sal ado Pe ona- Ga cel an a ,Juan F ancisco Rod  ıguez-Tes al a,c ,Ismael Hue as-Fe n andez b and Pablo Mi b,d 1 a G upo Neu odesa ollo y Psicosis, Ins i u o de Biomedicina de Se illa (IBIS), Hospi al Uni e si a io Vi gen del Rocio/CSIC/Uni e sidad de Se illa/UGC Salud Men al HVR, b Unidad de T as o nos del Mo imien o, Se icio de Neu olog ıa y Neu o isiolog ıa Cl ınica, Ins i u o de Biomedicina de Se illa, Hospi al Uni e si a io Vi gen del Roc ıo/CSIC/Uni e sidad de Se illa, c Depa amen o de Pe sonalidad, E aluaci on y T a amien os Psicol ogicos, Facul ad de Psicolog ıa, Uni e sidad de Se illa, Se ille, d en o de In es igaci on Biom edica en Red sob e En e medades Neu odegene a i as (CIBERNED), Spain 2 doi: 10.1111/bdi.12382 Key wo ds: co ical inhibi ion – GABA ecep o s – longi udinal s udy – mo o co ex – sho -in e al in aco ical inhibi ion – anscallosal inhibi ion Recei ed 30 June 2015, e ised and accep ed o publica ion 8 Feb ua y 2016 Co esponding au ho : Pablo Mi , M.D., Ph.D. Unidad de T as o nos del Mo imien o Se icio de Neu olog ıa y Neu o isiolog ıa Cl ınica Hospi al Uni e si a io Vi gen del Roc ıo A da. Manuel Siu o s/n Se illa 41013 Spain Fax: +34-955923101 E-mail: [email p o ec ed] † These au ho s con ibu ed equally o his wo k. Bipola diso de (BD) is a p e alen neu opsychi- a ic diso de cha ac e ized by pe iods o ele a ed mood (mania o hypomania) and dep ession. BD is associa ed wi h signi ican impai men in bo h he physical and men al quali y o li e, e en du ing eu hymic pe iods (1). One o he main symp oms o his diso de is an impai men in he inhibi ion o inapp op ia e ac ions 3and hough s, esul ing in 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 1 Bipola Diso de s 2016 ©2016 John Wiley & Sons A/S Published by John Wiley & Sons L d. BIPOLAR DISORDERS B D I 12382 Dispa ch: 8.3.16 CE: Sa a anan Jou nal Code Manusc ip No. No. o pages: 11 PE: Thanga aj o e amilia i y, impulsi i y, disinhibi ion, and de i- ci s in o e bea ing esponse supp ession (2). These beha io s a e pa icula ly d ama ic du ing mania. Howe e , indi iduals wi h BD also display inhibi- o y de ici s in dep ession and eu hymia (3–5), sug- ges ing ha inhibi o y de ici s could be pa o he co e symp oms o BD. I has been sugges ed ha inhibi o y de ici s in BD a e ela ed o impai men in co ical inhibi ion in which co ical c-aminobu- y ic acid (GABA) inhibi o y in e neu ons play a c ucial ole in inhibi ing he ac i i y o o he co i- cal neu ons (6). In his ega d, expe imen al e i- dence suppo s an in ol emen o he GABA sys em in he pa hophysiology o BD. Mo eo e , gene ic associa ion s udies ha e ound dis inc polymo phisms o GABA ecep o genes associ- a ed wi h BD (7, 8), and abno mali ies in he exp ession o mul iple GABA- ela ed p o eins ha e been epo ed in pos mo em s udies in pa ien s wi h BD (9, 10). In o ma ion abou he unc ionali y o he co i- cal GABAe gic sys em can be indi ec ly ob ained using ansc anial magne ic s imula ion (TMS). By using single- and pai ed-pulse TMS, i is possible o assess he ac i i y o he GABAA and GABAB ecep o s in he mo o co ex, as well as he physi- ological in e ac ions be ween exci a o y and inhi- bi o y ci cui s in human subjec s (11, 12). The sho -in e al co ical inhibi ion (SICI) pa adigm assesses he inhibi ion o he mo o ac i i y induced by a condi ioned magne ic s imulus ha is p esen ed sho ly (2–5 msec) be o e ano he s imu- lus ha is able o p o oke a measu able mo o e oked po en ial (MEP). This ype o inhibi ion has been ela ed o he ac i i y o GABAA ecep- o s (13). Addi ional measu es o co ical inhibi- ion include he co ical silen pe iod (CSP) and long-in e al in aco ical inhibi ion (LICI), which a e bo h measu es o long-las ing in aco ical inhibi ion and hough o be dependen on GABAB ecep o ac i i y (14, 15). Ano he co i- cal inhibi o y phenomenon, anscallosal inhibi- ion (TCI), can be assessed wi h TMS using ei he pai ed-pulse pa adigms o ia he ipsila e al silen pe iod. Al hough in e hemisphe ic co icoco ical inhibi o y mechanisms a e no ully unde s ood, hey equi e he in ac in eg i y o callosal ibe bundles, and p esumably also equi e a p ese ed in aco ical GABAe gic inhibi o y in e neu on sys em (16, 17). Few TMS pape s ha e s udied co ical inhibi- o y mechanisms in BD. Le inson e al. (18) demons a ed se e al co ical inhibi o y de ici s consis ing o dec eased SICI, TCI, and CSP, which sugges s impai ed co ical GABAe gic ecep o ac i i y in pa ien s wi h BD. By con as , LICI seems o be main ained in pa ien s wi h BD, as obse ed in a combined TMS–elec oen- cephalog aphy s udy (19). These p e ious s udies ha e in es iga ed he ai and s a e ma ke s o BD by compa ing co ical inhibi ion in eu hymic pa ien s wi h BD wi h con ol subjec s in a c oss- sec ional case-con ol design. The e is he e o e he possibili y o in e subjec di e ences a ec ing he a iabili y in co ical exci abili y, and i is di i- cul o in es iga e he unde lying co ical inhibi- o y de ici s o ai and s a e abno mali ies and mood swi ching using such a design. Subs ance abuse is a majo como bidi y in pa ien s wi h BD, which may ha e p o ound clinical implica ions (20, 21). Subs ances such as cannabis, cocaine, o nico ine ha e been shown o al e co ical inhibi- ion bo h in he non-psychia ic (22) and psychi- a ic popula ion (23). Likewise, co ical exci abili y and inhibi ion is also modula ed by commonly used d ugs o ea BD, including mood s abilize s, benzodiazepines, and an ipsy- cho ic agen s. The e o e, como bid subs ance abuse o cu en pha macological ea men should be aken in o accoun when assessing co i- cal exci abili y and inhibi ion in BD. To add ess hese issues, we conduc ed a longi udinal s udy wi h a TMS design o assess pa ien s du ing he acu e manic phase (BD manic) and emission om his acu e phase. We also assessed he associ- a ion be ween co ical inhibi o y measu es, se e - i y o psychia ic symp oms, and li e ime and cu en d ug use. Me hods Pa icipan s A o al o 19 igh -handed pa ien s wi h BD pa ic- ipa ed in he s udy ( en pa ien s had i s -episode BD; 12 males; mean age: 35.5 yea s; mean age a onse : 30.4 yea s). Pa ien s we e ec ui ed o e a wo-yea a pe iod om he Inpa ien Psychia ic Uni o he Hospi al Uni e si a io Vi gen del Roc ıo Men al Heal h Depa men . Eligible pa ien s me he diagnos ic c i e ia o bipola I diso de , acco ding o DSM-IV classi ica ion s an- da ds. The pa icipan s we e equi ed o ha e a Young Mania Ra ing Scale (YMRS) (24) sco e ≥18. We excluded pa ien s: (i) <18 o >65 yea s o age; (ii) wi h a his o y o b ain auma o neu o- logical disease; and (iii) i hey had unde gone elec- ocon ulsi e he apy in he p e ious 12 mon hs. All pa ien s displayed acu e mania wi h a YMRS >21 and none displayed majo dep essi e symp- oms. In i s -episode BD, diagnosis was con i med a e one yea o ollow-up. 2 Ruiz-Veguilla e al. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 A heal hy con ol g oup consis ing o 28 heal hy subjec s was ec ui ed by ad e isemen , in addi- ion o wo d-o -mou h eques s om s a in he esea ch uni . Pa ien and con ol g oups we e ma ched ia ec ui men by age, gende , and social/occupa ional class. The heal hy con ols exhibi ed no pas o p esen psychia ic o neu o- logical diso de s and had no posi i e amily his o y o psychia ic diso de s. O he exclusion c i e ia o his g oup we e li e ime subs ance dependence, subs ance abuse du ing he p e ious mon h, can- nabis abuse du ing he p e ious mon h, in ellec ual disabili y, demen ia, and neu ological illnesses. Six een pa ien s ag eed o pa icipa e in he ol- low-up assessmen and h ee wi hd ew du ing he cou se o he s udy ( wo o hem mo ed o o he ci ies a e discha ge and one no longe wished o pa icipa e u he in he s udy). All pa ien s we e eu hymic a he ollow-up assessmen , as de ined by a YMRS sco e <7 and a Hamil on Dep ession Ra ing Scale (HDRS) (25) sco e <10. Du ing he one-yea ollow-up, one pa ien displayed symp- oms consis en wi h schizoa ec i e diso de , so was excluded om he analysis. S udy design This was a longi udinal and na u alis ic s udy ha consis ed o comp ehensi e psychopa hological and TMS assessmen s a wo poin s: baseline and a h ee-mon h ollow-up. Baseline assessmen s we e comple ed in he second week o hospi al admission. Th ee mon hs a e hospi al discha ge, pa ien s we e in i ed o pa icipa e in a second assessmen . Heal hy subjec s unde wen a unique TMS assess- men . The s udy p o ocol was app o ed by he local e hics commi ee, and all p ocedu es ollowed we e in acco dance wi h ins i u ional guidelines. W i en in o med consen was ob ained om all subjec s. Following sc eening o exclusion c i e ia, all pa icipan s unde wen a de ailed examina ion by expe ienced psychia is s o assess cu en mood s a e. BD was diagnosed using he S uc u ed Clin- ical In e iew o DSM-IV Diso de s, esea che e sion wi h psycho ic sc een (26). Mood was also a ed using he YMRS and he HDRS. Psycho ic symp oms we e assessed using he Posi i e and Nega i e Symp oms Scale (PANSS) (27). T ained in e iewe s adminis e ed he PANSS as pa o a s uc u ed clinical in e iew and sco ed i ems on a scale om 1 (asymp oma ic) o 7 (ex emely symp- oma ic). I ems o he PANSS we e g ouped using Wallwo k e al.’s i e- ac o (posi i e, nega i e, diso ganized/conc e e, exci ed, and dep essed) model (28). Highe sco es in hese ac o s imply a highe se e i y o psycho ic symp oms. TMS assessmen s we e pe o med be ween 3:00 p.m. and 5:00 p.m. Doses o he a ypical an ipsy- cho ic agen s we e ans o med o chlo p omazine equi alen s (29). Du ing bo h assessmen s, benzo- diazepines we e wi hd awn a leas 15 hou s p io o he TMS s udy. Blood samples we e d awn o de e mina ion o se um li hium and alp oic acid le els du ing clinical examina ions. E ec i e se um li hium le els we e de ined as concen a- ions be ween 0.6 mEq/L and 1.0 mEq/L, while e ec i e alp oic acid le els we e de ined as be ween 50 mg/L and 100 mg/L. Pa e ns o can- nabis and cocaine abuse we e assessed using he L-sec ion o he Composi e In e na ional Diag- nos ic In e iew (CIDI) (30). We in e iewed sub- jec s on li e ime d ug use his o y and pa e ns o d ug use du ing he p e ious 12 mon hs. U ine oxicology sc eening (TOX/See; Bio-Rad 4, USA) was pe o med in all pa ien s o con i m o ule ou cu en exposu e o ha m ul d ugs. Nico ine dependence was assessed in bo h g oups using he Fage s € om es o nico ine dependence (31). A sco e ≥4 by combining i ems 1 and 4 o his es was used o iden i y subjec s wi h high nico ine dependence (32). Elec omyog aphy (EMG) eco dings Subjec s we e sea ed in a com o able chai and su ace EMG eco dings we e aken a he i s do sal in e osseous (FDI) muscles on he side con ala e al o he s imula ed co ex wi h Ag–AgCl su ace elec odes using a belly– endon mon age. EMG signals we e ampli ied (1,0009) and band-pass il e ed (bandwid h 20 Hz o 2 kHz) using a Digi ime D360 ampli ie (Dig- i ime 5, UK), acqui ed a a sampling a e o 5 kHz h ough a CED 1401 labo a o y in e ace (Camb idge Elec onic Design, Camb idge, UK) and s o ed on a compu e . The EMG aces we e analyzed using cus omized SIGNAL so wa e e sion 4 6. S imula ion p o ocols Single- and pai ed-pulse TMS o he p ima y mo o co ex we e applied using Mags im 200 magne ic s imula o s (Mags im Company Limi ed, Whi land, UK). The s imula o s we e igge ed h ough he SIGNAL so wa e and CED 1401 boa d. Pa ien s and heal hy subjec s we e es ed on he le hemisphe e in a TMS session ha included SICI, in aco ical acili a ion (ICF), LICI, and CSP. Fo hese s udies, he magne ic s imula o s we e connec ed o a s anda d igu e-o -eigh coil wi h an ex e nal diame e o 70 mm (peak mag- 3 Mo o co ical disinhibi ion in BD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 ne ic ield 2.2T). The coil was held angen ially o he skull wi h he handle poin ing backwa ds and la e ally a an angle o ~45 deg ees o he sagi al plane in o de o gene a e a pos e io –an e io cu - en in he b ain. A BiS im module (Mags im Company Limi ed) was used o in e connec s imu- la o s when pai ed-pulse TMS p o ocols we e pe - o med. SICI and ICF we e assessed o e he con ala e al FDI muscle in a simila way o a p e- iously desc ibed pai ed-pulse pa adigm (33), whe eby a sub h eshold s imulus is used o condi- ion he mo o ou pu o a sup a h eshold s imu- lus, depending on he ime in e al be ween hem. The in ensi y o he condi ioning s imulus (CS) was 80% o he ac i e mo o h eshold (AMT), de ined as he minimum in ensi y ha elici ed a ep oducible MEP o a leas 200 lV in he oni- cally con ac ing FDI muscle in a leas i e ou o en consecu i e ials, while subjec s we e con ac - ing a app oxima ely 20% o hei maximum ol- un a y con ac ion. A cons an le el o muscle con ac ion was achie ed by he use o isual eed- back. The in ensi y o he es s imulus (TEST) was adjus ed o elici an MEP o app oxima ely 1 mV. SICI and ICF we e assessed in he same expe i- men al block. SICI was assessed a es a in e - s imulus in e als (ISIs) be ween he CS and a TEST o 2 msec and 3 msec, and ICF a ISIs o 10 msec and 12 msec. The p esen a ion o TEST and CS we e andomly in e mingled 7wi hin he SICI–ICF expe imen al block. Fo assessmen o LICI, he in ensi y o he CS was 120% o he es - ing mo o h eshold (RMT) o he con ala e al FDI muscle. RMT was de ined as he minimum in ensi y ha e oked a peak- o-peak MEP o 50 lV in a leas i e ou o en consecu i e ials in he elaxed eco ded muscle (34). LICI was assessed a es a ISIs o 100 msec, 150 msec, and 250 msec. Ten MEPs we e collec ed o each ISI and o he TEST. The p esen a ion o TEST and CS s imuli we e andomly 8in e mingled wi hin he LICI expe imen al blocks. Fo assessmen o CSP, 15 single TMS pulses we e applied a an in ensi y o 120% RMT, while pa ien s p o ided a cons an con ac ion o he FDI muscle a 20% o hei maximum olun a y con ac ion, assis ed by isual eedback. TCI was demons a ed using he dual- pulse pa adigm desc ibed by Fe be e al. (17). B ie ly, wo igu e-o -eigh coils (wi h an ou e diame e o each hal -wing o 40 mm) we e used wi h a TEST applied o he le mo o co ex and a CS applied o he igh mo o co ex. Wi h hese coils, RMT was again ob ained a bo h hemi- sphe es 9. The TEST s imulus was se a an in ensi y ha , when gi en alone, would e oke an EMG esponse o 1 mV peak- o-peak ampli ude. The CS was applied a 120% o he RMT. TCI was es ed a ou condi ioning es in e als (7, 10, 45, and 75 msec), gi en in a andom o de , wi h a o al o en sweeps o 10each condi ion. S a is ical analysis S uden ’s - es s o independen measu es and Fishe ’s exac es we e used o in es iga e di e - ences be ween pa ien s and heal hy con ols on demog aphic a iables (i.e., age, gende , smoking a iables, and d ug abuse his o y a iables). Fish- e ’s exac es and pai ed-sample - es s we e used o assess changes in clinical a iables a he ol- low-up assessmen . The independen and epea ed-measu es - es s we e used o de e mine di e ences in mo o h esholds and CSP. Two sep- a a e wo-way ixed-e ec s analyses o a iance (ANOVAs) we e conduc ed o he e alua ion o pai ed-pulse TMS p o ocols. To de ec co ical inhibi o y de ici s du ing mania, TMS da a col- lec ed om pa ien s in his episode we e compa ed wi h hose ob ained om heal hy subjec s. This analysis was ca ied ou wi h ISI as wi hin-subjec ac o and G oup (heal hy e sus manic g oups) as be ween-subjec ac o . Po en ial con ounde s (use o cannabis, cocaine, obacco, and benzodi- azepines) we e included in a seconda y model analysis i a signi ican g oup di e ence was seen in he p ima y analysis. To assess he longi udinal e ec s o mood episodes on co ical inhibi ion, a epea ed-measu e ANOVA wi h wo wi hin-sub- jec ac o s, episode (manic and eu hymia) and ISI, was used. Mauchly’s es assessed sphe ici y and he G eenhouse–Geisse co ec ion was used o non-sphe ical da a. Signi ican main e ec s and in e ac ions in ANOVA we e ollowed by pos -hoc independen o pai ed - es s wi h Bon- e oni co ec ion. The ela ionship be ween clini- cal symp om se e i y and TMS measu es in he BD g oup was examined using he Pea son p o- duc -momen co ela ion coe icien . A Bon e oni co ec ion was applied o he co ela ion analysis, se ing he signi icance cu -o a p <0.001. A powe analysis using he Gpowe compu e p o- g am 11(35) indica ed ha , gi en ou sample size, we would expec o de ec mode a e e ec sizes ( =0.25), powe ed, 86%, a a signi ican le el o p<0.05. Resul s Demog aphic da a and a his o y o subs ance use bo h in pa ien s wi h BD and heal hy subjec s a e displayed in Table 1. Mos pa ien s wi h BD we e cu en ciga e e smoke s, al hough nico ine depen- 4 Ruiz-Veguilla e al. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 dence measu es did no di e s a is ically be ween pa ien s and con ols. Ten pa ien s epo ed a his- o y o cannabis use, while daily cannabis use was asce ained in six o hem. A posi i e u ine es o cannabis was ound in nine o hem. In all o hese cases, ime since he las cannabis exposu e exceeded one week. Fou pa ien s epo ed a his o y o cocaine use. All o hem, excep one, epo ed no cocaine use in he p e ious yea . A posi i e u ine es o cocaine was p esen in one pa ien , who epo ed hei las cocaine consump ion o ha e aken place in he p e ious mon h. Table 2 shows changes in clinical and pha ma- cological a iables in pa ien s du ing mania and eu hymia. O e all, a signi ican imp o emen in manic, posi i e, and gene al psycho ic symp oms was obse ed (all p <0.001) in he ollow-up assessmen . Ten ou o he 15 pa ien s sco ed ze o in he YMRS. We also de ec ed a signi ican inc ease in he se e i y o dep essi e symp oms in he ollow-up assessmen . Fu he mo e, a signi i- can dec ease in cannabis consump ion (i.e., es s posi i e o cannabis) was obse ed du ing ollow-up. Mo o h esholds The da a on mo o h esholds a e shown in Table 3. No signi ican di e ences could be demons a ed in RMT in pa ien s wi h BD (p =0.75 o pa ien s in he manic phase and p=0.12 o eu hymic pa ien s). The e we e no s a- is ically signi ican di e ences be ween heal hy con ols and pa ien s wi h BD in he o he mo o h esholds ob ained o igh FDI muscle (all p>0.37). Likewise, no signi ican di e ences in he mo o h esholds we e ound be ween pa ien s du ing he manic episode and du ing eu hymia (all p>0.86). Co ical inhibi ion du ing he manic episode SICI and ICF. A signi ican main e ec o ISI (F 3,132 =49.6, p <0.001) and ISI 9g oup in e ac ion (F 3,44 =6.6, p <0.001) we e ob ained. Table 1. Pa icipan demog aphics and subs ance use da a Bipola diso de (n =19) Heal hy con ols (n =28) p- alue Gende , male/ emale 12/7 18/10 0.937 Age, yea s, mean SD 35.5 11.4 33.1 7.0 0.434 Cu en ciga e e smoke , n 13 5 <0.001 FTND sco e, mean SD 3.19 0.9 2.92 0.4 0.168 High nico ine dependence, n 6 2 0.065 Cannabis use his o y, n 10 0 <0.001 Posi i e u ine es , n 9 0 N/A Daily smoke , n 6 –N/A Weekly smoke , n 4 – Time since las join , days, mean SD 17.3 27.6 –N/A Age a onse , yea s, mean SD 19.4 4.1 –N/A Du a ion o use, yea s, mean SD 9.5 5.1 –N/A Cocaine use his o y, n 4 0 0.022 Posi i e u ine es , n 1 0 N/A Daily cocaine use , n 2 –N/A 3–4 days a week, n 1 –N/A 1–2 days a week, n 1 –N/A Age a onse , yea s, mean SD 24.2 6.7 –N/A Du a ion o use, yea s, mean SD 7.4 5.5 –N/A O he d ugs, n 0 0 N/A FTND =Fage s € om es o nico ine dependence; N/A =no applicable; SD =s anda d de ia ion. Table 2. Clinical cha ac e is ics o pa ien s who comple ed he ollow- up s udy Manic episode (n =15) Eu hymic episode (n =15) p- alue a Mood s abilize , n Li hium 7 3 0.428 Sodium alp oa e 8 8 Se um li hium le els, mmol/L, mean (SD) 0.61 (0.08) 0.72 (0.21) 0.336 Se um alp oa e le els, mg/L, mean (SD) 59.6 (23.4) 49.2 (11.2) 0.346 An ipsycho ic agen s, n 15 13 0.483 An ipsycho ic agen doses, chlo p omazine equi alen s, mean (SD) 649.8 (339.3) 595.8 (380.1) 0.392 Benzodiazepines, n 6 3 0.427 Cu en ciga e e smoke , n 9 9 1.000 Cannabis use, n 8 2 0.020 YMRS sco e, mean (SD) 35.1 (7.1) 1.5 (2.8) <0.001 HDRS sco e, mean (SD) 1.2 (1.6) 2.9 (2.5) 0.039 PANSS sco e, mean (SD) Posi i e ac o 3 (1.1) 1.1 (0.3) <0.001 Nega i e ac o 1.3 (0.9) 1.4 (0.7) 0.309 Diso ganized/ Conc e e ac o 3.4 (0.8) 1.3 (0.5) <0.001 Exci ed ac o 4.1 (0.9) 1.2 (0.3) <0.001 Dep essed ac o 1.6 (0.1) 1.6 (0.7) 0.770 HDRS =Hamil on Dep ession Ra ing Scale; PANSS =Posi i e and Nega i e Synd ome Scale; SD =s anda d de ia ion; YMRS =Young Mania Ra ing Scale. a Fishe ’s exac es o pai ed-samples - es s. 5 Mo o co ical disinhibi ion in BD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 Pos -hoc analysis e ealed ha pa ien s wi h BD du ing mania showed less SICI a ISIs 2 msec (p <0.001) and 3 msec (p <0.05), while no di e - ences we e ound in ICF (p >0.2) (Fig. 1A). A e - aged ac oss bo h inhibi o y ISIs (2 msec and 3 msec), pa ien s du ing mania demons a ed 33.2% less inhibi ion compa ed wi h heal hy sub- jec s ( 44 =3.3, p =0.002). The e we e no di e - ences be ween g oups when ICF measu es we e pooled (p =0.203). Di e ences in SICI measu es emained s a is ically signi ican a e con olling o cu en obacco smoking (es ima ed ma ginal means o SICI in he BD g oup =0.78, heal hy g oup =0.36, p =0.01). We also assessed he e ec o cu en o pas cannabis use on SICI measu es in pa ien s wi h BD. In he mul i a ia e model, he e was a signi ican e ec o ISI (p <0.001), while nei he he ISI 9cu en cannabis use (p =0.571) no cu en cannabis use main ac o (p =0.947) eached signi ican le els. Simila esul s we e ob ained a e including he a iables equency o cu en cannabis smoking (in e ac ion wi h ISI, p =0.287; main e ec , p =0.655), ime since las join (p =0.166), age a cannabis use onse (p =0.639), o du a ion o cannabis use (p =0.734) on SICI 12. His o y o cocaine use did no show a sig- ni ican main e ec (p =0.482) o in e ac ion wi h ISI (p =0.117). Pa ien s on benzodiazepine ea - men showed enhanced SICI a bo h 2 msec and 3 msec (bo h p <0.05, as compa ed wi h pa ien s wi hou benzodiazepines). Long-in e al co ical 14inhibi ion. A main e ec o ISI (F 3,135 =17.28, p <0.001) was ound, indica - ing a signi ican MEP educ ion when a condi- ioned s imulus was applied bo h o con ols and pa ien s (p <0.001). Howe e , no signi ican main e ec o g oup (F 1,47 =0.022, p =0.883) o he in e ac ion ISI 9g oup (F 3,141 =0.449, p =0.718) was ound o his p o ocol. CSP. The du a ion o CSP did no di e be ween pa ien s wi h mania and heal hy con ols (mean CSP du a ion manic pa ien s =110.3 msec; con- ols =116.09 msec; 45 =0.708, p =0.483). TCI. One pa ien could no comple e his p o o- col. Signi ican main e ec s o ISI (F 3,129 =22.1, p<0.001) and g oup (F 1,43 =7.8, p =0.008) we e ob ained, while he g oup 9ISI in e ac ion was no signi ican (F 4,43 =1.05, p =0.359) (Fig. 1B). Planned - es s showed dec eased TCI in manic pa ien s a ISIs 7 msec (p =0.021) and 10 msec (p =0.018). A e a e aging ac oss all sho -in e - al in e hemisphe ic inhibi o y ISIs (i.e., 7 msec and 10 msec), pa ien s wi h BD displayed less inhi- bi ion han heal hy con ols (17.5%, p =0.033). An in e ac ion be ween cu en obacco smoking and g oup was also de ec ed (F 1,44 =4.33, p=0.044). Pos -hoc analysis e ealed ha smoke Table 3. TMS in ensi y (% o maximum s imula o ou pu ) used in he TMS assessmen s o pa ien s wi h bipola diso de and heal hy sub- jec s Heal hy con ols Manic episode Eu hymic episode Ac i e mo o h eshold ( igh FDI muscle) 31.8 1.3 32.1 1.7 32.2 1.8 Res ing mo o h eshold ( igh FDI muscle) 40.8 1.3 42.5 2.1 41.2 2.5 In ensi y MEP 1 mV ( igh FDI muscle) 49.4 1.9 51.4 2.5 51.2 2.6 Res ing mo o h eshold ( igh FDI muscle) a 43.0 1.8 42.7 2.2 44.9 2.3 Res ing mo o h eshold (le FDI muscle) a 41.7 1.4 43.9 3.2 43.3 2.6 Values epo ed as mean s anda d de ia ion. FDI = i s do - sal in e osseous; MEP =mo o e oked po en ial; TMS = an- sc anial magne ic s imula ion. a Th esholds ob ained using 40-mm coils di ec ly connec ed o he s imula o s. AB TEST 2 msec 3 msec 10 msec 12 msec 0.0 0.5 1.0 1.5 MEP ampli ude No malized by TEST ampli ude *** * TEST 7 msec 10 msec 45 msec 75 msec 0.0 0.5 1.0 MEP ampli ude No malized by TEST ampli ude * * Fig. 1. (A) Sho -in e al in aco ical inhibi ion and in aco ical acili a ion in pa ien s wi h bipola diso de and heal hy subjec s. (B) T anscallosal inhibi ion in pa ien s wi h bipola diso de and heal hy subjec s as a unc ion o he in e s imulus in e al be ween he condi ioning and he es s imuli. Mean s anda d e o om pa ien s du ing he manic episode (black ba s, n =19) and heal hy con ols (g ay ba s, n =28). Values <1 indica e inhibi ion. *p<0.05, ***p13 <0.001: independen - es s wi h Bon e oni co - ec ion o mul iple compa isons. MEP =mo o e oked po en ial. 6 Ruiz-Veguilla e al. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 pa ien s wi h BD displayed TCI le els ha we e compa able wi h hose ob ained in he heal hy g oup (p =0.441), while TCI le els in non-smoke pa ien s wi h BD we e signi ican ly educed (p =0.018, as compa ed wi h heal hy subjec s). Cu en obacco smoking did no a ec TCI le els in heal hy subjec s (p =0.690). A cu en o pas his o y o cannabis use did no show he main e ec (p =0.221) o in e ac ion wi h ISI (p =0.368). Likewise, equency o cu en cannabis smoking did no show an e ec on TCI esul s (in- e ac ion wi h ISI, p =0.637; main e ec , p=0.257). O he a iables ha did no in luence TCI le els in pa ien s wi h BD included ime since las join (p =0.993), age a cannabis use onse (p =0.999), and du a ion o cannabis use (p =0.837). A his o y o cocaine use did no show a signi ican main e ec (p =0.419) o in e ac ion wi h ISI (p =0.611). Nei he he main e ec o benzodiazepine (p =0.287) no in e ac ion o ben- zodiazepine wi h TCI ISIs we e signi ican (p =0.596). Co ical inhibi ion du ing eu hymia SICI and ICF. A signi ican main e ec o ISI (F 3,36 =10.06, p <0.001), bu no episode (F 1,12 =0.007, p =0.935) o in e ac ion episode 9 ISI (F 3,36 =0.056, p =0.784), was ound in he ol- low-up s udy, indica ing no di e ences be ween pa ien s du ing manic and eu hymic phases a any e alua ed condi ion 15 (Fig. 2A). Cu en obacco smoking in he model did no s a is ically al e hese esul s (p =0.575). Cannabis, cocaine, and benzodiazepine e ec s we e no analyzed a he ol- low-up assessmen owing o he small numbe o pa ien s unde hese condi ions. Long-in e al co ical inhibi ion 17 .No signi ican main e ec o episode (F 1,13 =0.196, p =0.665) o in e ac ion episode 9ISI (F 2,26 =1.764, p=0.191) was ound in he ollow-up s udy. Silen pe iod. No signi ican di e ences we e ound be ween pa ien s in manic and eu hymic phases ( 14 =1.009, p =0.331). TCI. Analysis o di e ences be ween he manic episode and he ollow-up assessmen e ealed a signi ican main ac o ISI (F 3,39 =7.94, p<0.001), whe eas nei he main ac o episode (F 1,13 =1.132, p =0.307) no he in e ac ion be ween episode and ISI (F 3,36 =0.623, p =0.604) we e signi ican (Fig. 2B). A signi ican main e ec o obacco was also ound in he mul i a ia e anal- ysis (F 1,13 =4.86, p =0.046). Cu en obacco smoke s showed an inc eased mean TCI compa ed wi h non-smoke s. Rela ionship be ween se e i y o psychia ic symp oms and measu es o co ical inhibi ion SICI and TCI le els (a e aged ac oss all inhibi o y ISIs) co ela ed posi i ely bo h in manic episode and eu hymia. Pa ien s wi h a high sco e in he PANSS exci ed ac o du ing he manic episode displayed less SICI and TCI du ing bo h he manic episode and eu hymia. PANSS diso ganized ac o co ela ed posi i ely wi h he posi i e and nega i e ac o s o he same scale. Mo eo e , he PANSS dep essed ac o showed a nega i e co ela ion wi h he HDS 18du ing mania. These co ela ions did no each signi icance a e he Bon e oni co - ec ion. The ull co ela ion ma ix is p esen ed in Supplemen a y Table 1. We also conduc ed a co ela ion analysis wi h YMRS i em 2, which assesses inc eased mo o ac i i y and ene gy. The sco e in his i em co e- la ed nega i ely wi h he deg ee o co ical inhibi- ion in bo h manic and eu hymic episodes bu TEST 2 msec 3 msec 10 msec 12 msec 0.0 0.5 1.0 1.5 A MEP ampli ude No malized by TEST ampli ude Tes 7 msec 10 msec 45 msec 75 msec 0.0 0.5 1.0 B MEP ampli ude No malized by TEST ampli ude Fig. 2. (A) Sho -in e al in aco ical inhibi ion and in aco ical acili a ion in pa ien s wi h bipola diso de du ing he manic epi- sode (g ay ba s) and eu hymia (whi e ba s 16 ). (B) T anscallosal inhibi ion in pa ien s du ing he manic episode and eu hymia. Fi een pa ien s we e included in his ollow-up s udy. Values <1 indica e inhibi ion. No signi ican di e ence in co ical inhibi ion be ween manic episode and eu hymia was ound. Pai ed sample - es s wi h Bon e oni co ec ion o mul iple compa ison. MEP =mo o e oked po en ial. 7 Mo o co ical disinhibi ion in BD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 hese did no each signi icance a e he Bon e - oni co ec ion (SICI mania: ho =0.632, p=0.006; SICI eu hymia: ho =0.562, p=0.046; TCI mania: ho =0.526, p =0.065; TCI eu hymia: ho =0.409, p =0.103). Discussion We ound ha pa ien s wi h BD du ing mania dis- play speci ic co ical inhibi o y de ici s –namely, al e ed SICI and TCI. The deg ee o co ical inhibi- ion was simila du ing mania and eu hymia, sug- ges ing ha co ical inhibi o y de ici s cons i u e a ai ma ke o BD. In an explo a o y co ela ion analysis, se e al associa ions be ween psychopa ho- logical symp om se e i y and co ical inhibi o y measu es we e de ec ed. I should be highligh ed ha pa ien s who sco ed high on i ems assessing mo o hype ac i i y and goal-di ec ed beha io (i.e., PANSS exci ed ac o i ems 19 , YMRS i em 2) du ing mania displayed highe co ical SICI and TCI de ici s bo h in manic episodes and eu hymia. Se e al e o s ha e been made o dis inguish which cogni i e, emo ional, and mo i a ional al e a ions o BD a e mood-s a e dependen o ai ma ke s by examining pa ien s ac oss a ious mood s a es, including eu hymia (36, 37). How- e e , ai - and s a e-dependen co ical inhibi o y de ici s ha e no been ex ensi ely s udied, despi e he ac ha hese de ici s ha e been linked o impulsi eness, a co e symp om o BD (21, 38, 39). Indeed, clinical scales o assess BD usually include mo o ac i i y i ems. Thus, i may seem easonable o e alua e ai - and s a e-dependen co ical exci abili y/inhibi ion as an indica o o he neu- opa hophysiology unde lying mo o de ici s in mood diso de s. Ou esul s, which con i m and u he ex end p e ious obse a ions in eu hymic pa ien s (18, 19), include se e al new obse a ions on mania ha p o ide a mo e comple e unde - s anding o mo o inhibi ion de ici s in BD. In pa - icula , he longi udinal design used in ou s udy empowe ed he gene alizabili y o ou da a o he comple e cou se o BD, no limi ed o a speci ic ype o episode. Al e ed co ical inhibi ion has been obse ed in se e e psycho ic and mood diso de s, in spi e o he concomi an 20 ele a ed pe iphe al and b ain GABA le els (40, 41). The e o e, a plausible hypo hesis o he mechanism in ol ed in co ical inhibi o y de ici s seen in BD would in ol e al e - a ions in he GABA ecep o sys em. Pha maco- logical s udies sugges he in ol emen o GABAA ecep o s in SICI assessed in he p ima y mo o co ex. The e o e, and in ag eemen wi h Le inson e al. (18), de ici s in SICI could e idence an impai men in as inhibi o y co ical p ocesses likely o in ol e GABAA ecep o s (42). We also ound ha benzodiazepines modula ed he SICI le el in pa ien s wi h BD du ing mania – ha is, pa ien s aking benzodiazepines du ing he manic episode displayed simila SICI le els o heal hy con ols, while pa ien s wi hou his ea men dis- played signi ican ly educed SICI le els. The e ec o benzodiazepines on SICI could no be es ed du ing eu hymia as only h ee pa ien s we e aking benzodiazepines a ollow-up. The enhancing e ec o benzodiazepines on SICI is consis en ly epo ed in he pha maco-TMS li e a u e (43) and could unde lie he indings in BD epo ed he e. The e- o e, i is possible ha he SICI educ ion obse ed in he manic phase could ha e been e en highe han ha ob ained in he cu en s udy i pa ien s had no been on benzodiazepine ea men . This possibili y does no make ou main indings oid as manic pa ien s s ill show de ici s in SICI, and his does no change du ing eu hymia. In addi ion o de ici s ound in SICI, in e hemi- sphe ic inhibi ion a sho ISIs (7 msec and 10 msec) was also ound o be abno mal bo h in manic and eu hymic episodes. Al hough he exac mechanisms media ing TCI emain unde in es i- ga ion, he e is a en a i e consensus on he in ol emen o anscallosal glu ama e gic pa h- ways linking wi h py amidal ac neu ons h ough GABAe gic in e neu ons (44). In humans, TCI is a obus phenomenon and occu s o e a wide ange o ISIs (6–50 msec). Howe e , eme ging e idence sugges s ha TCI elici ed a sho ISIs is media ed by di e en mechanisms han ha elici ed a longe in e als (45). Animal s udies ha e sugges ed ha he ea ly las ing o m o TCI is media ed by GABAA ecep o s (46). Ne e heless, da a in humans a e inconclusi e (47, 48). Ano he po en- ial ac o ha may unde lie TCI de ici s in BD ega ds ana omical al e a ions in he co pus callo- sum, such as he educed callosal wid h docu- men ed in his popula ion (49, 50). Subs ance use was no con olled in ou s udy. The e o e, he e ec o hese a iables on co ical inhibi o y measu es was analyzed sepa a ely in mul i a ia e analyses. Consis en wi h o he s ud- ies, pa ien s wi h BD displayed high a es o cu - en and li e ime cannabis, cocaine, and obacco use (20, 51). The analysis o cannabis and cocaine use showed no e ec o hese subs ances on TMS measu es in BD du ing mania. We also obse ed a signi ican dec ease in he numbe o pa ien s wi h- ou cannabis and cocaine use du ing ollow-up (e.g., eigh pa ien s epo ed ha hey had s opped smoking cannabis in ecen mon hs, which was u - he con i med by he u ine d ug es , and no 8 Ruiz-Veguilla e al. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56