Connectivity between brain areas is increasingly being suggested to play a key role in neurophysiology. However, measuring connectivity is a challenging task. Multiple methods exist, but many are indirect and probe different aspects of connectivity, such as anatomical pathways connecting two areas versus functional co-activation of brain areas. Disorders in which structural damage to white matter connections occurs can provide new insights into the physiology of connections between distant brain areas, and can be useful testing grounds for whether ‘disconnectivity’ leads to pathology. One such disorder is chronic obstructive pulmonary disease (COPD), in which reduced lung capacity is accompanied by brain abnormalities and additional sequelae, including muscle weakness and cognitive impairment. While some studies have explored brain abnormalities in COPD, few have focused on connectivity changes and their underlying neurophysiological mechanisms. Given previous evidence of callosal damage in COPD patients, in a recent issue of The Journal of Physiology Cabibel and colleagues (2020) hypothesized that connectivity between the primary motor cortices may be disrupted. To test this, they used transcranial magnetic stimulation (TMS) to measure interhemispheric connectivity in a case–control design comparing 22 COPD patients with 21 age and sex-matched controls. In partial agreement with the authors’ hypothesis, COPD patients and controls differed in two measures of connectivity, the ipsilateral silent period (iSP) and inter-hemispheric inhibition (IHI; Fig. 1, right panel), but not in a third, cross-activation (Fig. 1, centre-right panel), which was quantified using motor evoked potential (MEP) and short intracortical inhibition (SICI) measured during contraction of the ipsilateral leg. The authors find that ISP and IHI measures of interhemispheric connectivity are lower in COPD compared to controls (rightmost panel), but Mmax (leftmost panel), Hmax (centre-left panel), MEP and SICI measures (centre-right panel), across both rest and muscle contraction, are similar between COPD patients and controls. Pathways are colour-coded based on their involvement in each measurement; the limb depicted in blue represents the side where EMG was collected, with the outcome being M-waves, H-waves, MEPs or changes in MEPs in each panel, respectively. All limbs are lower-limbs, with the exception of IHI representing hand first dorsal interosseus muscles. What do these results reveal about the role of connectivity in COPD? iSP, IHI and cross-activation are all indirect measures of inter-hemispheric connectivity. The fact that iSP and IHI, but not cross-activation, are impaired in COPD, might seem counter-intuitive. Cabibel et al. speculate that this may be due to reorganization of inter-hemispheric pathways in COPD, but other competing interpretations cannot be ruled out. One is that cross-activation, iSP and IHI may have different degrees of sensitivity to different aspects of inter-hemispheric connectivity. Inter-hemispheric connectivity refers to the strength of connections between the two hemispheres but it is not a unitary phenomenon, and rather encompasses multiple structural and functional aspects of connections between hemispheres, including multiple fibre populations and neuromodulatory systems. While most measures of connectivity are likely inter-related, different metrics may be most sensitive to, for example, excitatory vs. inhibitory interhemispheric modulation. Another possibility is that all inter-hemispheric measures considered may engage the same mechanism, but with different degrees of sensitivity. Although the effect size of cross-activation differences between COPD patients and controls was extremely small (η2p< 0.01), it could be that iSP is a more sensitive measure, or that cross-activation is noisier. In this scenario, cross-activation differences between COPD and controls may be present, but only detectable at larger sample sizes. Moreover, neither iSP nor IHI are specific measures of interhemispheric connectivity, and both are sensitive to variability in excitability at the cortical, spinal and peripheral levels. Cabibel et al. take these into consideration using a variety of measures. The lack of changes in MEP and SICI measurements at rest make COPD-related changes in corticospinal excitability and intracortical inhibition unlikely. The authors also fail to detect differences in spinal and peripheral excitability using Mmax and Hmax amplitude. While these findings suggest local excitability of the primary motor cortex and spine does not drive the iSP/IHI differences observed in this sample, they cannot exclude that corticospinal excitability plays a role in clinical manifestations of COPD. COPD is a heterogeneous disorder, with varying severity of muscle weakness and cognitive symptoms which is likely a reflection of different neurophysiological mechanisms. Because the authors only studied chronic COPD patients, and the study was not powered to detect differences between subtypes, these results may not extend to COPD patients with severe muscle weakness, who might be expected to have lower cortical excitability (Alexandre et al. 2020). Regardless of their interpretation, the study findings highlight the translational potential of TMS-based measures of connectivity, which have been predominantly used in healthy subjects. However, a remaining obstacle to the clinical translatability of these measures is the lack of knowledge on their test–retest reliability. IHI is known to have low reliability even in healthy adults (De Gennaro et al. 2003). While iSP is known to have moderate reliability (ICC > 0.6) in healthy adults, its reliability in older subjects and clinical populations is still untested (Fleming & Newham, 2017). Therefore, additional investigations are needed to gain a deeper understanding of the true clinical potential of TMS-based measures of connectivity. Taken together, this study suggests that disruption of brain connectivity may play a key role in COPD. However, it is unclear whether compensatory changes take place in response to these reductions in connectivity, and whether local cortical and spinal alterations also contribute to clinical manifestations of COPD subtypes. What mechanisms underlie altered connectivity between primary motor cortices in COPD? Cabibel et al. find that iSP as well as both short IHI (SIHI; 10 ms interstimulus interval) and long IHI (LIHI; 40 ms interstimulus interval) are decreased in COPD. These measures are thought to engage a combination of GABAA and GABAB interneuronal circuits, suggesting multiple inhibitory mechanisms may be impaired in COPD. Since interhemispheric inhibition, particularly IHI, is thought to be driven by excitatory corpus callosal fibres acting on local inhibitory circuits, interpreting the interhemispheric changes in light of local intracortical inhibition may further our understanding of the observed ‘dysconnectivity’ between primary motor cortices. One putative inhibitory mechanism that the authors consider is GABAA-mediated inhibition, measured using SICI. The authors find no evidence for differences between controls and COPD patients in levels of SICI inhibition. However, different TMS protocols reflect different aspects of GABAA activity, and a potential role for GABAA-ergic transmission in the SIHI deficits observed in COPD cannot be excluded without using triple pulse techniques and measurements in different muscle contraction states. Additionally, to measure SICI, the authors use a single conditioning stimulus at a fixed intensity, which has the disadvantage that each participant is measured at a different point in their inhibitory recruitment curve. To further dissect the role of GABAA inhibition changes in COPD, future studies may want to use a wider range of conditioning pulse intensities, as well as multiple inter-pulse intervals, which are known to engage separate GABAA-ergic mechanisms. Another local inhibitory mechanism potentially contributing to the results observed by Cabibel et al. is GABAB-mediated inhibition. While little is known about the role of GABAB inhibition in COPD at rest, two previous COPD studies found increases in the cortico-silent period (CSP), which is measured during muscle contraction and is thought to be sensitive to GABAB-ergic activity in the later part (>100 ms) (Alexandre et al. 2020 in chronic COPD and Mohamed-Hussein et al. 2007 in acute exacerbation patients). As both CSP and LIHI likely engage GABAB mechanisms (Schnitzler et al. 1996), increased CSP may seem to be at odds with the decreased LIHI/iSP observed by Cabibel et al. However, triple pulse TMS studies (Schnitzler et al. 1996) have shown that activation of interhemispheric pathways reduces CSP, and a COPD-related loss of inter-hemispheric activation may lead to local disinhibition, thus explaining the increase in CSP. In summary, we lack a clear understanding of interactions between inter-hemispheric pathways and local GABAA inhibition in COPD. The emerging picture about GABAB inhibition suggests that COPD related loss of inter-hemispheric excitatory pathways leads to changes in the activity of local inhibitory circuits that they synapse on. Such changes may have implications beyond motor weakness, since these local inhibitory circuits are known to also play a key role in skill learning. Previous studies investigating connectivity in COPD have used neuroimaging-based connectivity metrics, such as diffusion-weighted imaging, and resting-state functional magnetic resonance imaging, which are often influenced by vascular factors. As COPD patients are known to have abnormal vasculature (such as increased arterial stiffness and impaired vasodilatation), excluding vascular effects in neuroimaging studies of COPD is challenging. Cabibel et al. take a novel approach by using TMS to measure connectivity in COPD. While TMS-based measures of connectivity have lower spatial resolution compared to neuroimaging-based ones, focusing on physiological aspects of connectivity is an apt methodological approach to study connectivity in COPD, as it is not confounded by vascular factors and thus allows a more direct study of connectivity between brain areas. Moreover, the choice of complementing lower-limb iSP with upper-limb IHI provides strong evidence for disruption in connectivity between motor cortices across the somatotopic motor map. However, TMS also has methodological disadvantages compared to neuroimaging. For example, COPD patients do not exhibit focal lesions in the corpus callosum, but rather distributed, global abnormalities in both grey and white matter, which may be measured with structural neuroimaging techniques. Without neuroimaging markers of callosal damage available in the sample of COPD patients used here, it is difficult to exclude that lesions outside the callosum may be influencing the TMS-based readouts. Combining neuroimaging and TMS-based markers of callosal integrity may allow further insights into the relationship between white matter damage and connectivity underlying mechanisms of muscle weakness in COPD. Harnessing a combination of indirect metrics can be a powerful approach to studying the role of brain connectivity in disease. Since connectivity alterations are likely to play a role in COPD, neuroimaging, together with other TMS-based metrics, could help in better understanding the nature and underlying mechanisms of the connectivity changes observed by Cabibel and colleagues. This kind of multimodal approach will be crucial to disentangling how connectivity shapes heterogeneous clinical presentations not only in COPD subtypes, but also in other disorders involving structural brain damage. None. All authors have read and approved the final version of this manuscript and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed. A.L. is funded by the Wellcome Trust (grant number 109062/Z/15/Z). We would like to thank Charlotte J. Stagg and Melanie Fleming for their feedback on earlier versions of this commentary.