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

Depósito de Investigación de la Universidad de Sevilla https://idus.us.es/ This is the peer reviewed version of the following article: Ruiz-Veguilla, M., Martín- Rodríguez, J. F., Palomar, F. J., Porcacchia, P., Álvarez de Toledo, P., Perona-Garcelán, S., Rodríguez-Testal, J. F., Huertas-Fernández, I., & Mir, P. (2016). Trait- and statedependent cortical inhibitory deficits in bipolar disorder. Bipolar Disorders, 18(3), 261– 271, which has been published in final form at https://doi.org/10.1111/bdi.12382 . This article may be used for non-commercial purposes in accordance with Wiley Terms and Conditions for Use of Self-Archived Versions. This article may not be enhanced, enriched or otherwise transformed into a derivative work, without express permission from Wiley or by statutory rights under applicable legislation. Copyright notices must not be removed, obscured or modified. The article must be linked to Wiley’s version of record on Wiley Online Library and any embedding, framing or otherwise making available the article or pages thereof by third parties from platforms, services and websites other than Wiley Online Library must be prohibited. Original Article Trait- and state-dependent cortical inhibitory deficits in bipolar disorder Ruiz-Veguilla M, Mart ın-Rodr ıguez JF, Palomar FJ, Porcacchia P,  Alvarez de Toledo P, Perona-Garcel an S, Rodr ıguez-Testal JF, Huertas-Fern andez I, Mir P. Trait- and state-dependent cortical inhibitory deficits in bipolar disorder. Bipolar Disord 2016: 00: 000–000. ©2016 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd. Objectives: Euthymic patients with bipolar disorder (BD) have deficits in cortical inhibition. However, whether cortical inhibitory deficits are trait- or state-dependent impairments is not yet known and their relationship with psychiatric symptoms is not yet understood. In the present study, we examined trait- and state-dependent cortical inhibitory deficits and evaluated the potential clinical significance of these deficits. Methods: Nineteen patients with bipolar I disorder were evaluated using the paired-pulse transcranial stimulation protocol, which assessed cortical inhibition during an acute manic episode. Cortical inhibition measures were compared with those obtained in 28 demographically matched healthy controls. A follow-up assessment was performed in 15 of these patients three months later, when there was remission from their mood and psychotic symptoms. The association between cortical inhibitory measures and severity of psychiatric symptoms was also studied. Results: During mania, patients showed decreased short-interval intracortical and transcallosal inhibition, as well as a normal cortical silent period and long-interval cortical inhibition. These findings were the same during euthymia. Symptoms associated with motor hyperactivity were correlated negatively with the degree of cortical inhibition. These correlations were not significant when a Bonferroni correction was applied. Conclusions: The present longitudinal study showed cortical inhibitory deficits in patients with BD, and supports the hypothesis that cortical inhibitory deficits in BD are trait dependent. Further research is necessary to confirm the clinical significance of these deficits. Miguel Ruiz-Veguilla a,† ,Juan Francisco Mart ın-Rodr ıguez b, † , Francisco J Palomar b ,Paolo Porcacchia b ,Paloma  Alvarez de Toledo b ,Salvador Perona- Garcel an a ,Juan Francisco Rodr ıguez-Testal a,c ,Ismael Huertas-Fern andez b and Pablo Mir b,d 1 a Grupo Neurodesarrollo y Psicosis, Instituto de Biomedicina de Sevilla (IBIS), Hospital Universitario Virgen del Rocio/CSIC/Universidad de Sevilla/UGC Salud Mental HVR, b Unidad de Trastornos del Movimiento, Servicio de Neurolog ıa y Neurofisiolog ıa Cl ınica, Instituto de Biomedicina de Sevilla, Hospital Universitario Virgen del Roc ıo/CSIC/Universidad de Sevilla, c Departamento de Personalidad, Evaluaci on y Tratamientos Psicol ogicos, Facultad de Psicolog ıa, Universidad de Sevilla, Seville, d entro de Investigaci on Biom edica en Red sobre Enfermedades Neurodegenerativas (CIBERNED), Spain 2 doi: 10.1111/bdi.12382 Key words: cortical inhibition – GABA receptors – longitudinal study – motor cortex – short-interval intracortical inhibition – transcallosal inhibition Received 30 June 2015, revised and accepted for publication 8 February 2016 Corresponding author: Pablo Mir, M.D., Ph.D. Unidad de Trastornos del Movimiento Servicio de Neurolog ıa y Neurofisiolog ıa Cl ınica Hospital Universitario Virgen del Roc ıo Avda. Manuel Siurot s/n Sevilla 41013 Spain Fax: +34-955923101 E-mail: [email protected] † These authors contributed equally to this work. Bipolar disorder (BD) is a prevalent neuropsychiatric disorder characterized by periods of elevated mood (mania or hypomania) and depression. BD is associated with significant impairment in both the physical and mental quality of life, even during euthymic periods (1). One of the main symptoms of this disorder is an impairment in the inhibition of inappropriate actions 3and thoughts, resulting 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 Bipolar Disorders 2016 ©2016 John Wiley & Sons A/S Published by John Wiley & Sons Ltd. BIPOLAR DISORDERS B D I 12382 Dispatch: 8.3.16 CE: Saravanan Journal Code Manuscript No. No. of pages: 11 PE: Thangaraj overfamiliarity, impulsivity, disinhibition, and deficits in overbearing response suppression (2). These behaviors are particularly dramatic during mania. However, individuals with BD also display inhibitory deficits in depression and euthymia (3–5), suggesting that inhibitory deficits could be part of the core symptoms of BD. It has been suggested that inhibitory deficits in BD are related to impairment in cortical inhibition in which cortical c-aminobu- tyric acid (GABA) inhibitory interneurons play a crucial role in inhibiting the activity of other cortical neurons (6). In this regard, experimental evidence supports an involvement of the GABA system in the pathophysiology of BD. Moreover, genetic association studies have found distinct polymorphisms of GABA receptor genes associated with BD (7, 8), and abnormalities in the expression of multiple GABA-related proteins have been reported in postmortem studies in patients with BD (9, 10). Information about the functionality of the cortical GABAergic system can be indirectly obtained using transcranial magnetic stimulation (TMS). By using single- and paired-pulse TMS, it is possible to assess the activity of the GABAA and GABAB receptors in the motor cortex, as well as the physiological interactions between excitatory and inhibitory circuits in human subjects (11, 12). The short-interval cortical inhibition (SICI) paradigm assesses the inhibition of the motor activity induced by a conditioned magnetic stimulus that is presented shortly (2–5 msec) before another stimulus that is able to provoke a measurable motor evoked potential (MEP). This type of inhibition has been related to the activity of GABAA receptors (13). Additional measures of cortical inhibition include the cortical silent period (CSP) and long-interval intracortical inhibition (LICI), which are both measures of long-lasting intracortical inhibition and thought to be dependent on GABAB receptor activity (14, 15). Another cortical inhibitory phenomenon, transcallosal inhibition (TCI), can be assessed with TMS using either paired-pulse paradigms or via the ipsilateral silent period. Although interhemispheric corticocortical inhibitory mechanisms are not fully understood, they require the intact integrity of callosal fiber bundles, and presumably also require a preserved intracortical GABAergic inhibitory interneuron system (16, 17). Few TMS papers have studied cortical inhibitory mechanisms in BD. Levinson et al. (18) demonstrated several cortical inhibitory deficits consisting of decreased SICI, TCI, and CSP, which suggests impaired cortical GABAergic receptor activity in patients with BD. By contrast, LICI seems to be maintained in patients with BD, as observed in a combined TMS–electroencephalography study (19). These previous studies have investigated the trait and state markers of BD by comparing cortical inhibition in euthymic patients with BD with control subjects in a crosssectional case-control design. There is therefore the possibility of intersubject differences affecting the variability in cortical excitability, and it is difficult to investigate the underlying cortical inhibitory deficits for trait and state abnormalities and mood switching using such a design. Substance abuse is a major comorbidity in patients with BD, which may have profound clinical implications (20, 21). Substances such as cannabis, cocaine, or nicotine have been shown to alter cortical inhibition both in the non-psychiatric (22) and psychiatric population (23). Likewise, cortical excitability and inhibition is also modulated by commonly used drugs to treat BD, including mood stabilizers, benzodiazepines, and antipsychotic agents. Therefore, comorbid substance abuse or current pharmacological treatment should be taken into account when assessing cortical excitability and inhibition in BD. To address these issues, we conducted a longitudinal study with a TMS design to assess patients during the acute manic phase (BD manic) and remission from this acute phase. We also assessed the association between cortical inhibitory measures, severity of psychiatric symptoms, and lifetime and current drug use. Methods Participants A total of 19 right-handed patients with BD participated in the study (ten patients had first-episode BD; 12 males; mean age: 35.5 years; mean age at onset: 30.4 years). Patients were recruited over a two-year a period from the Inpatient Psychiatric Unit of the Hospital Universitario Virgen del Roc ıo Mental Health Department. Eligible patients met the diagnostic criteria for bipolar I disorder, according to DSM-IV classification standards. The participants were required to have a Young Mania Rating Scale (YMRS) (24) score ≥18. We excluded patients: (i) <18 or >65 years of age; (ii) with a history of brain trauma or neurological disease; and (iii) if they had undergone electroconvulsive therapy in the previous 12 months. All patients displayed acute mania with a YMRS >21 and none displayed major depressive symptoms. In first-episode BD, diagnosis was confirmed after one year of follow-up. 2 Ruiz-Veguilla et 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 healthy control group consisting of 28 healthy subjects was recruited by advertisement, in addition to word-of-mouth requests from staff in the research unit. Patient and control groups were matched via recruitment by age, gender, and social/occupational class. The healthy controls exhibited no past or present psychiatric or neurological disorders and had no positive family history of psychiatric disorders. Other exclusion criteria for this group were lifetime substance dependence, substance abuse during the previous month, cannabis abuse during the previous month, intellectual disability, dementia, and neurological illnesses. Sixteen patients agreed to participate in the follow-up assessment and three withdrew during the course of the study (two of them moved to other cities after discharge and one no longer wished to participate further in the study). All patients were euthymic at the follow-up assessment, as defined by a YMRS score <7 and a Hamilton Depression Rating Scale (HDRS) (25) score <10. During the one-year follow-up, one patient displayed symptoms consistent with schizoaffective disorder, so was excluded from the analysis. Study design This was a longitudinal and naturalistic study that consisted of comprehensive psychopathological and TMS assessments at two points: baseline and at three-month follow-up. Baseline assessments were completed in the second week of hospital admission. Three months after hospital discharge, patients were invited to participate in a second assessment. Healthy subjects underwent a unique TMS assessment. The study protocol was approved by the local ethics committee, and all procedures followed were in accordance with institutional guidelines. Written informed consent was obtained from all subjects. Following screening for exclusion criteria, all participants underwent a detailed examination by experienced psychiatrists to assess current mood state. BD was diagnosed using the Structured Clinical Interview for DSM-IV Disorders, researcher version with psychotic screen (26). Mood was also rated using the YMRS and the HDRS. Psychotic symptoms were assessed using the Positive and Negative Symptoms Scale (PANSS) (27). Trained interviewers administered the PANSS as part of a structured clinical interview and scored items on a scale from 1 (asymptomatic) to 7 (extremely symptomatic). Items of the PANSS were grouped using Wallwork et al.’s five-factor (positive, negative, disorganized/concrete, excited, and depressed) model (28). Higher scores in these factors imply a higher severity of psychotic symptoms. TMS assessments were performed between 3:00 p.m. and 5:00 p.m. Doses of the atypical antipsychotic agents were transformed to chlorpromazine equivalents (29). During both assessments, benzodiazepines were withdrawn at least 15 hours prior to the TMS study. Blood samples were drawn for determination of serum lithium and valproic acid levels during clinical examinations. Effective serum lithium levels were defined as concentrations between 0.6 mEq/L and 1.0 mEq/L, while effective valproic acid levels were defined as between 50 mg/L and 100 mg/L. Patterns of cannabis and cocaine abuse were assessed using the L-section of the Composite International Diagnostic Interview (CIDI) (30). We interviewed subjects on lifetime drug use history and patterns of drug use during the previous 12 months. Urine toxicology screening (TOX/See; Bio-Rad 4, USA) was performed in all patients to confirm or rule out current exposure to harmful drugs. Nicotine dependence was assessed in both groups using the Fagerstr€ om test for nicotine dependence (31). A score ≥4 by combining items 1 and 4 of this test was used to identify subjects with high nicotine dependence (32). Electromyography (EMG) recordings Subjects were seated in a comfortable chair and surface EMG recordings were taken at the first dorsal interosseous (FDI) muscles on the side contralateral to the stimulated cortex with Ag–AgCl surface electrodes using a belly–tendon montage. EMG signals were amplified (1,0009) and band-pass filtered (bandwidth 20 Hz to 2 kHz) using a Digitimer D360 amplifier (Digitimer 5, UK), acquired at a sampling rate of 5 kHz through a CED 1401 laboratory interface (Cambridge Electronic Design, Cambridge, UK) and stored on a computer. The EMG traces were analyzed using customized SIGNAL software version 4 6. Stimulation protocols Single- and paired-pulse TMS of the primary motor cortex were applied using Magstim 200 magnetic stimulators (Magstim Company Limited, Whitland, UK). The stimulators were triggered through the SIGNAL software and CED 1401 board. Patients and healthy subjects were tested on the left hemisphere in a TMS session that included SICI, intracortical facilitation (ICF), LICI, and CSP. For these studies, the magnetic stimulators were connected to a standard figure-of-eight coil with an external diameter of 70 mm (peak mag- 3 Motor cortical disinhibition 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 netic field 2.2T). The coil was held tangentially to the skull with the handle pointing backwards and laterally at an angle of ~45 degrees to the sagittal plane in order to generate a posterior–anterior current in the brain. A BiStim module (Magstim Company Limited) was used to interconnect stimulators when paired-pulse TMS protocols were performed. SICI and ICF were assessed over the contralateral FDI muscle in a similar way to a previously described paired-pulse paradigm (33), whereby a subthreshold stimulus is used to condition the motor output of a suprathreshold stimulus, depending on the time interval between them. The intensity of the conditioning stimulus (CS) was 80% of the active motor threshold (AMT), defined as the minimum intensity that elicited a reproducible MEP of at least 200 lV in the tonically contracting FDI muscle in at least five out of ten consecutive trials, while subjects were contracting at approximately 20% of their maximum voluntary contraction. A constant level of muscle contraction was achieved by the use of visual feedback. The intensity of the test stimulus (TEST) was adjusted to elicit an MEP of approximately 1 mV. SICI and ICF were assessed in the same experimental block. SICI was assessed at rest at interstimulus intervals (ISIs) between the CS and a TEST of 2 msec and 3 msec, and ICF at ISIs of 10 msec and 12 msec. The presentation of TEST and CS were randomly intermingled 7within the SICI–ICF experimental block. For assessment of LICI, the intensity of the CS was 120% of the resting motor threshold (RMT) of the contralateral FDI muscle. RMT was defined as the minimum intensity that evoked a peak-to-peak MEP of 50 lV in at least five out of ten consecutive trials in the relaxed recorded muscle (34). LICI was assessed at rest at ISIs of 100 msec, 150 msec, and 250 msec. Ten MEPs were collected for each ISI and for the TEST. The presentation of TEST and CS stimuli were randomly 8intermingled within the LICI experimental blocks. For assessment of CSP, 15 single TMS pulses were applied at an intensity of 120% RMT, while patients provided a constant contraction of the FDI muscle at 20% of their maximum voluntary contraction, assisted by visual feedback. TCI was demonstrated using the dualpulse paradigm described by Ferbert et al. (17). Briefly, two figure-of-eight coils (with an outer diameter of each half-wing of 40 mm) were used with a TEST applied to the left motor cortex and a CS applied to the right motor cortex. With these coils, RMT was again obtained at both hemispheres 9. The TEST stimulus was set at an intensity that, when given alone, would evoke an EMG response of 1 mV peak-to-peak amplitude. The CS was applied at 120% of the RMT. TCI was tested at four conditioning test intervals (7, 10, 45, and 75 msec), given in a random order, with a total of ten sweeps for 10each condition. Statistical analysis Student’s t-tests for independent measures and Fisher’s exact test were used to investigate differences between patients and healthy controls on demographic variables (i.e., age, gender, smoking variables, and drug abuse history variables). Fisher’s exact test and paired-sample t-tests were used to assess changes in clinical variables at the follow-up assessment. The independent and repeated-measures t-tests were used to determine differences in motor thresholds and CSP. Two separate two-way fixed-effects analyses of variance (ANOVAs) were conducted for the evaluation of paired-pulse TMS protocols. To detect cortical inhibitory deficits during mania, TMS data collected from patients in this episode were compared with those obtained from healthy subjects. This analysis was carried out with ISI as within-subject factor and Group (healthy versus manic groups) as between-subject factor. Potential confounders (use of cannabis, cocaine, tobacco, and benzodiazepines) were included in a secondary model analysis if a significant group difference was seen in the primary analysis. To assess the longitudinal effects of mood episodes on cortical inhibition, a repeated-measure ANOVA with two within-sub- ject factors, episode (manic and euthymia) and ISI, was used. Mauchly’s test assessed sphericity and the Greenhouse–Geisser correction was used for non-spherical data. Significant main effects and interactions in ANOVA were followed by post-hoc independent or paired t-tests with Bonferroni correction. The relationship between clinical symptom severity and TMS measures in the BD group was examined using the Pearson product-moment correlation coefficient. A Bonferroni correction was applied to the correlation analysis, setting the significance cut-off at p <0.001. A power analysis using the Gpower computer program 11(35) indicated that, given our sample size, we would expect to detect moderate effect sizes (f=0.25), powered, 86%, at a significant level of p<0.05. Results Demographic data and a history of substance use both in patients with BD and healthy subjects are displayed in Table 1. Most patients with BD were current cigarette smokers, although nicotine depen- 4 Ruiz-Veguilla et 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 measures did not differ statistically between patients and controls. Ten patients reported a history of cannabis use, while daily cannabis use was ascertained in six of them. A positive urine test for cannabis was found in nine of them. In all of these cases, time since the last cannabis exposure exceeded one week. Four patients reported a history of cocaine use. All of them, except one, reported no cocaine use in the previous year. A positive urine test for cocaine was present in one patient, who reported their last cocaine consumption to have taken place in the previous month. Table 2 shows changes in clinical and pharmacological variables in patients during mania and euthymia. Overall, a significant improvement in manic, positive, and general psychotic symptoms was observed (all p <0.001) in the follow-up assessment. Ten out of the 15 patients scored zero in the YMRS. We also detected a significant increase in the severity of depressive symptoms in the follow-up assessment. Furthermore, a significant decrease in cannabis consumption (i.e., tests positive for cannabis) was observed during follow-up. Motor thresholds The data on motor thresholds are shown in Table 3. No significant differences could be demonstrated in RMT in patients with BD (p =0.75 for patients in the manic phase and p=0.12 for euthymic patients). There were no statistically significant differences between healthy controls and patients with BD in the other motor thresholds obtained for right FDI muscle (all p>0.37). Likewise, no significant differences in the motor thresholds were found between patients during the manic episode and during euthymia (all p>0.86). Cortical inhibition during the manic episode SICI and ICF. A significant main effect of ISI (F 3,132 =49.6, p <0.001) and ISI 9group interaction (F 3,44 =6.6, p <0.001) were obtained. Table 1. Participant demographics and substance use data Bipolar disorder (n =19) Healthy controls (n =28) p-value Gender, male/female 12/7 18/10 0.937 Age, years, mean SD 35.5 11.4 33.1 7.0 0.434 Current cigarette smoker, n 13 5 <0.001 FTND score, mean SD 3.19 0.9 2.92 0.4 0.168 High nicotine dependence, n 6 2 0.065 Cannabis use history, n 10 0 <0.001 Positive urine test, n 9 0 N/A Daily smoker, n 6 –N/A Weekly smoker, n 4 – Time since last joint, days, mean SD 17.3 27.6 –N/A Age at onset, years, mean SD 19.4 4.1 –N/A Duration of use, years, mean SD 9.5 5.1 –N/A Cocaine use history, n 4 0 0.022 Positive urine test, n 1 0 N/A Daily cocaine user, n 2 –N/A 3–4 days a week, n 1 –N/A 1–2 days a week, n 1 –N/A Age at onset, years, mean SD 24.2 6.7 –N/A Duration of use, years, mean SD 7.4 5.5 –N/A Other drugs, n 0 0 N/A FTND =Fagerstr€ om test for nicotine dependence; N/A =not applicable; SD =standard deviation. Table 2. Clinical characteristics of patients who completed the followup study Manic episode (n =15) Euthymic episode (n =15) p-value a Mood stabilizer, n Lithium 7 3 0.428 Sodium valproate 8 8 Serum lithium levels, mmol/L, mean (SD) 0.61 (0.08) 0.72 (0.21) 0.336 Serum valproate levels, mg/L, mean (SD) 59.6 (23.4) 49.2 (11.2) 0.346 Antipsychotic agents, n 15 13 0.483 Antipsychotic agent doses, chlorpromazine equivalents, mean (SD) 649.8 (339.3) 595.8 (380.1) 0.392 Benzodiazepines, n 6 3 0.427 Current cigarette smoker, n 9 9 1.000 Cannabis use, n 8 2 0.020 YMRS score, mean (SD) 35.1 (7.1) 1.5 (2.8) <0.001 HDRS score, mean (SD) 1.2 (1.6) 2.9 (2.5) 0.039 PANSS score, mean (SD) Positive factor 3 (1.1) 1.1 (0.3) <0.001 Negative factor 1.3 (0.9) 1.4 (0.7) 0.309 Disorganized/ Concrete factor 3.4 (0.8) 1.3 (0.5) <0.001 Excited factor 4.1 (0.9) 1.2 (0.3) <0.001 Depressed factor 1.6 (0.1) 1.6 (0.7) 0.770 HDRS =Hamilton Depression Rating Scale; PANSS =Positive and Negative Syndrome Scale; SD =standard deviation; YMRS =Young Mania Rating Scale. a Fisher’s exact test or paired-samples t-tests. 5 Motor cortical disinhibition 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 Post-hoc analysis revealed that patients with BD during mania showed less SICI at ISIs 2 msec (p <0.001) and 3 msec (p <0.05), while no differences were found in ICF (p >0.2) (Fig. 1A). Averaged across both inhibitory ISIs (2 msec and 3 msec), patients during mania demonstrated 33.2% less inhibition compared with healthy subjects (t 44 =3.3, p =0.002). There were no differences between groups when ICF measures were pooled (p =0.203). Differences in SICI measures remained statistically significant after controlling for current tobacco smoking (estimated marginal means of SICI in the BD group =0.78, healthy group =0.36, p =0.01). We also assessed the effect of current or past cannabis use on SICI measures in patients with BD. In the multivariate model, there was a significant effect of ISI (p <0.001), while neither the ISI 9current cannabis use (p =0.571) nor current cannabis use main factor (p =0.947) reached significant levels. Similar results were obtained after including the variables frequency of current cannabis smoking (interaction with ISI, p =0.287; main effect, p =0.655), time since last joint (p =0.166), age at cannabis use onset (p =0.639), or duration of cannabis use (p =0.734) on SICI 12. History of cocaine use did not show a significant main effect (p =0.482) or interaction with ISI (p =0.117). Patients on benzodiazepine treatment showed enhanced SICI at both 2 msec and 3 msec (both p <0.05, as compared with patients without benzodiazepines). Long-interval cortical 14inhibition. A main effect of ISI (F 3,135 =17.28, p <0.001) was found, indicating a significant MEP reduction when a conditioned stimulus was applied both to controls and patients (p <0.001). However, no significant main effect of group (F 1,47 =0.022, p =0.883) or the interaction ISI 9group (F 3,141 =0.449, p =0.718) was found for this protocol. CSP. The duration of CSP did not differ between patients with mania and healthy controls (mean CSP duration manic patients =110.3 msec; controls =116.09 msec; t 45 =0.708, p =0.483). TCI. One patient could not complete this protocol. Significant main effects of ISI (F 3,129 =22.1, p<0.001) and group (F 1,43 =7.8, p =0.008) were obtained, while the group 9ISI interaction was not significant (F 4,43 =1.05, p =0.359) (Fig. 1B). Planned t-tests showed decreased TCI in manic patients at ISIs 7 msec (p =0.021) and 10 msec (p =0.018). After averaging across all short-inter- val interhemispheric inhibitory ISIs (i.e., 7 msec and 10 msec), patients with BD displayed less inhibition than healthy controls (17.5%, p =0.033). An interaction between current tobacco smoking and group was also detected (F 1,44 =4.33, p=0.044). Post-hoc analysis revealed that smoker Table 3. TMS intensity (% of maximum stimulator output) used in the TMS assessments of patients with bipolar disorder and healthy subjects Healthy controls Manic episode Euthymic episode Active motor threshold (right FDI muscle) 31.8 1.3 32.1 1.7 32.2 1.8 Resting motor threshold (right FDI muscle) 40.8 1.3 42.5 2.1 41.2 2.5 Intensity MEP 1 mV (right FDI muscle) 49.4 1.9 51.4 2.5 51.2 2.6 Resting motor threshold (right FDI muscle) a 43.0 1.8 42.7 2.2 44.9 2.3 Resting motor threshold (left FDI muscle) a 41.7 1.4 43.9 3.2 43.3 2.6 Values reported as mean standard deviation. FDI =first dorsal interosseous; MEP =motor evoked potential; TMS =transcranial magnetic stimulation. a Thresholds obtained using 40-mm coils directly connected to the stimulators. AB TEST 2 msec 3 msec 10 msec 12 msec 0.0 0.5 1.0 1.5 MEP amplitude Normalized by TEST amplitude *** * TEST 7 msec 10 msec 45 msec 75 msec 0.0 0.5 1.0 MEP amplitude Normalized by TEST amplitude * * Fig. 1. (A) Short-interval intracortical inhibition and intracortical facilitation in patients with bipolar disorder and healthy subjects. (B) Transcallosal inhibition in patients with bipolar disorder and healthy subjects as a function of the interstimulus interval between the conditioning and the test stimuli. Mean standard error from patients during the manic episode (black bars, n =19) and healthy controls (gray bars, n =28). Values <1 indicate inhibition. *p<0.05, ***p13 <0.001: independent t-tests with Bonferroni correction for multiple comparisons. MEP =motor evoked potential. 6 Ruiz-Veguilla et 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 patients with BD displayed TCI levels that were comparable with those obtained in the healthy group (p =0.441), while TCI levels in non-smoker patients with BD were significantly reduced (p =0.018, as compared with healthy subjects). Current tobacco smoking did not affect TCI levels in healthy subjects (p =0.690). A current or past history of cannabis use did not show the main effect (p =0.221) or interaction with ISI (p =0.368). Likewise, frequency of current cannabis smoking did not show an effect on TCI results (interaction with ISI, p =0.637; main effect, p=0.257). Other variables that did not influence TCI levels in patients with BD included time since last joint (p =0.993), age at cannabis use onset (p =0.999), and duration of cannabis use (p =0.837). A history of cocaine use did not show a significant main effect (p =0.419) or interaction with ISI (p =0.611). Neither the main effect of benzodiazepine (p =0.287) nor interaction of benzodiazepine with TCI ISIs were significant (p =0.596). Cortical inhibition during euthymia SICI and ICF. A significant main effect of ISI (F 3,36 =10.06, p <0.001), but not episode (F 1,12 =0.007, p =0.935) or interaction episode 9 ISI (F 3,36 =0.056, p =0.784), was found in the follow-up study, indicating no differences between patients during manic and euthymic phases at any evaluated condition 15 (Fig. 2A). Current tobacco smoking in the model did not statistically alter these results (p =0.575). Cannabis, cocaine, and benzodiazepine effects were not analyzed at the follow-up assessment owing to the small number of patients under these conditions. Long-interval cortical inhibition 17 .No significant main effect of episode (F 1,13 =0.196, p =0.665) or interaction episode 9ISI (F 2,26 =1.764, p=0.191) was found in the follow-up study. Silent period. No significant differences were found between patients in manic and euthymic phases (t 14 =1.009, p =0.331). TCI. Analysis of differences between the manic episode and the follow-up assessment revealed a significant main factor ISI (F 3,39 =7.94, p<0.001), whereas neither main factor episode (F 1,13 =1.132, p =0.307) nor the interaction between episode and ISI (F 3,36 =0.623, p =0.604) were significant (Fig. 2B). A significant main effect of tobacco was also found in the multivariate analysis (F 1,13 =4.86, p =0.046). Current tobacco smokers showed an increased mean TCI compared with non-smokers. Relationship between severity of psychiatric symptoms and measures of cortical inhibition SICI and TCI levels (averaged across all inhibitory ISIs) correlated positively both in manic episode and euthymia. Patients with a high score in the PANSS excited factor during the manic episode displayed less SICI and TCI during both the manic episode and euthymia. PANSS disorganized factor correlated positively with the positive and negative factors of the same scale. Moreover, the PANSS depressed factor showed a negative correlation with the HDS 18during mania. These correlations did not reach significance after the Bonferroni correction. The full correlation matrix is presented in Supplementary Table 1. We also conducted a correlation analysis with YMRS item 2, which assesses increased motor activity and energy. The score in this item correlated negatively with the degree of cortical inhibition in both manic and euthymic episodes but TEST 2 msec 3 msec 10 msec 12 msec 0.0 0.5 1.0 1.5 A MEP amplitude Normalized by TEST amplitude Test 7 msec 10 msec 45 msec 75 msec 0.0 0.5 1.0 B MEP amplitude Normalized by TEST amplitude Fig. 2. (A) Short-interval intracortical inhibition and intracortical facilitation in patients with bipolar disorder during the manic episode (gray bars) and euthymia (white bars 16 ). (B) Transcallosal inhibition in patients during the manic episode and euthymia. Fifteen patients were included in this follow-up study. Values <1 indicate inhibition. No significant difference in cortical inhibition between manic episode and euthymia was found. Paired sample t-tests with Bonferroni correction for multiple comparison. MEP =motor evoked potential. 7 Motor cortical disinhibition 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 these did not reach significance after the Bonferroni correction (SICI mania: rho =0.632, p=0.006; SICI euthymia: rho =0.562, p=0.046; TCI mania: rho =0.526, p =0.065; TCI euthymia: rho =0.409, p =0.103). Discussion We found that patients with BD during mania display specific cortical inhibitory deficits –namely, altered SICI and TCI. The degree of cortical inhibition was similar during mania and euthymia, suggesting that cortical inhibitory deficits constitute a trait marker of BD. In an exploratory correlation analysis, several associations between psychopathological symptom severity and cortical inhibitory measures were detected. It should be highlighted that patients who scored high on items assessing motor hyperactivity and goal-directed behavior (i.e., PANSS excited factor items 19 , YMRS item 2) during mania displayed higher cortical SICI and TCI deficits both in manic episodes and euthymia. Several efforts have been made to distinguish which cognitive, emotional, and motivational alterations of BD are mood-state dependent or trait markers by examining patients across various mood states, including euthymia (36, 37). However, trait- and state-dependent cortical inhibitory deficits have not been extensively studied, despite the fact that these deficits have been linked to impulsiveness, a core symptom of BD (21, 38, 39). Indeed, clinical scales to assess BD usually include motor activity items. Thus, it may seem reasonable to evaluate trait- and state-dependent cortical excitability/inhibition as an indicator of the neuropathophysiology underlying motor deficits in mood disorders. Our results, which confirm and further extend previous observations in euthymic patients (18, 19), include several new observations on mania that provide a more complete understanding of motor inhibition deficits in BD. In particular, the longitudinal design used in our study empowered the generalizability of our data to the complete course of BD, not limited to a specific type of episode. Altered cortical inhibition has been observed in severe psychotic and mood disorders, in spite of the concomitant 20 elevated peripheral and brain GABA levels (40, 41). Therefore, a plausible hypothesis for the mechanism involved in cortical inhibitory deficits seen in BD would involve alterations in the GABA receptor system. Pharmacological studies suggest the involvement of GABAA receptors in SICI assessed in the primary motor cortex. Therefore, and in agreement with Levinson et al. (18), deficits in SICI could evidence an impairment in fast inhibitory cortical processes likely to involve GABAA receptors (42). We also found that benzodiazepines modulated the SICI level in patients with BD during mania –that is, patients taking benzodiazepines during the manic episode displayed similar SICI levels to healthy controls, while patients without this treatment displayed significantly reduced SICI levels. The effect of benzodiazepines on SICI could not be tested during euthymia as only three patients were taking benzodiazepines at follow-up. The enhancing effect of benzodiazepines on SICI is consistently reported in the pharmaco-TMS literature (43) and could underlie the findings in BD reported here. Therefore, it is possible that the SICI reduction observed in the manic phase could have been even higher than that obtained in the current study if patients had not been on benzodiazepine treatment. This possibility does not make our main findings void as manic patients still show deficits in SICI, and this does not change during euthymia. In addition to deficits found in SICI, interhemispheric inhibition at short ISIs (7 msec and 10 msec) was also found to be abnormal both in manic and euthymic episodes. Although the exact mechanisms mediating TCI remain under investigation, there is a tentative consensus on the involvement of transcallosal glutamatergic pathways linking with pyramidal tract neurons through GABAergic interneurons (44). In humans, TCI is a robust phenomenon and occurs over a wide range of ISIs (6–50 msec). However, emerging evidence suggests that TCI elicited at short ISIs is mediated by different mechanisms than that elicited at longer intervals (45). Animal studies have suggested that the early lasting form of TCI is mediated by GABAA receptors (46). Nevertheless, data in humans are inconclusive (47, 48). Another potential factor that may underlie TCI deficits in BD regards anatomical alterations in the corpus callosum, such as the reduced callosal width documented in this population (49, 50). Substance use was not controlled in our study. Therefore, the effect of these variables on cortical inhibitory measures was analyzed separately in multivariate analyses. Consistent with other studies, patients with BD displayed high rates of current and lifetime cannabis, cocaine, and tobacco use (20, 51). The analysis of cannabis and cocaine use showed no effect of these substances on TMS measures in BD during mania. We also observed a significant decrease in the number of patients without cannabis and cocaine use during follow-up (e.g., eight patients reported that they had stopped smoking cannabis in recent months, which was further confirmed by the urine drug test, and no 8 Ruiz-Veguilla et 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