Research Article: When Breathing Interferes with Cognition: Experimental Inspiratory Loading Alters Timed Up-and-Go Test in Normal Humans

Date Published: March 15, 2016

Publisher: Public Library of Science

Author(s): Marie-Cécile Nierat, Suela Demiri, Elise Dupuis-Lozeron, Gilles Allali, Capucine Morélot-Panzini, Thomas Similowski, Dan Adler, Joseph Najbauer.


Human breathing stems from automatic brainstem neural processes. It can also be operated by cortico-subcortical networks, especially when breathing becomes uncomfortable because of external or internal inspiratory loads. How the “irruption of breathing into consciousness” interacts with cognition remains unclear, but a case report in a patient with defective automatic breathing (Ondine’s curse syndrome) has shown that there was a cognitive cost of breathing when the respiratory cortical networks were engaged. In a pilot study of putative breathing-cognition interactions, the present study relied on a randomized design to test the hypothesis that experimentally loaded breathing in 28 young healthy subjects would have a negative impact on cognition as tested by “timed up-and-go” test (TUG) and its imagery version (iTUG). Progressive inspiratory threshold loading resulted in slower TUG and iTUG performance. Participants consistently imagined themselves faster than they actually were. However, progressive inspiratory loading slowed iTUG more than TUG, a finding that is unexpected with regard to the known effects of dual tasking on TUG and iTUG (slower TUG but stable iTUG). Insofar as the cortical networks engaged in response to inspiratory loading are also activated during complex locomotor tasks requiring cognitive inputs, we infer that competition for cortical resources may account for the breathing-cognition interference that is evidenced here.

Partial Text

In healthy humans, normal breathing stems from automatic brainstem neural processes and does not give rise to conscious perception: it does not engage motor or sensory cortical resources. Breathing can however be operated by cortico-subcortical networks under certain circumstances, like voluntary respiratory movements or during speech [1]. Cortically driven breathing has also been described in reaction to changes in the mechanical properties of the respiratory system, namely when breathing becomes difficult [2, 3]. Externally applied inspiratory and expiratory constraints give rise to respiratory-related motor cortical activities that are associated with an augmented neural drive to breathe [2–4]. The corresponding network involves the supplementary motor cortex, with emphasis on the supplementary motor area (SMA) [5]. Similar cortical activations have been reported in patients suffering from chronic respiratory insufficiency due to respiratory muscle weakness in the contexte of amyotrophic lateral sclerosis [6] and from the obstructive sleep apnea syndrome in which upper airway abnormalities generate an “intrinsic” inspiratory load [7]. Finally, a respiratory-related cortical activity exists during resting breathing in patients with defective respiratory automatism (Ondine’s curse syndrome)[8]. In one such patient, cognitive performances were better under mechanical ventilation than during to spontaneous breathing. This cognitive improvement was concomitant with a reduction in overall cortical activity, changes in brain functional connectivity (stronger connectivity between brainstem and frontal lobe during spontaneous breathing than during mechanical ventilation), and restoration of the default mode network that is associated with self-consciousness, mind-wandering, creativity and introspection [9]. This was interpreted as the result of “competition for cortical resources”, in the general frame of dual tasking interferences. It could thus be postulated that respiratory diseases involving a respiratory-related motor cortical activity could be associated with executive defects through such a mechanism, and irrespective of their impact on blood oxygen and carbon dioxide. Of note, inspiratory loads give rise to respiratory discomfort and negative emotions (namely “dyspnea”). This is associated with increased metabolic activities within the limbic cortex [10] and with deactivation of the default mode network [5]. This irruption of “breathing into consciousness” could also be a called upon to explain a negative impact of dyspnea on cognitive functions, by analogy with pain [11]

Twenty-five participants fully completed the experiment whereas 3 did not perform the breath-holding condition (added to the protocole afterwards)(see S1 Dataset). Median TUG time during quiet breathing was 8.11 s [IQR: 7.10–9.24]. It was 9.01 s [IQR: 6.99–10.14]) during 60% ITL, and 8.70 s [IQR: 8.16–9.33] during breath-holding. Median iTUG times during quiet breathing was 5.40 s [IQR: 4.40–6.51]. It was 6.85 s [IQR: 5.36–8.26]) during 60% ITL, and 6.45 s [IQR: 5.87–7.58] during breath-holding (Fig 3). TUG and iTUG increases appeared linear insofar as only the linear component of the polynomial contrast was significant (p<0.0001): only the linear part of the contrast was therefore kept in the analyses. There was a statistically significant main effect of task (p<0.001) and a significant interaction between task and inspiratory loading (p = 0.022), suggesting that inspiratory loading has a greater effect on iTUG time than on TUG time. The median value of the observed absolute TUG-iTUG difference decreased between spontaneous breathing (2.60 s [IQR: 2.01–3.47]) and 60% ITL (2.08 s [IQR: 0.72–3.12]). It was 2.28 s [IQR: 1.39–2.62] for breath-holding. Of notice, the randomization procedure was efficient, producing 17 different sequences among the 28 subjects (10 of the sequences pertained to one subject only, 4 to 2 subjects, 2 to 3 subjects, and 1 to 4 subjects; there was no statistically significant "order" effect on the results. This study shows that inspiratory threshold loading, known to elicit dyspnea and to engage respiratory-related cortical networks [19] interferes with the control of locomotion. Indeed, TUG and iTUG were slower under inspiratory loading than during quiet breathing. In addition, participants consistently imagined themselves being faster to perform the task than they actually were, but the TUG-iTUG difference decreased with increasing inspiratory loading (greater impact of inspiratory loading on iTUG), which is unexpected.   Source: