Le bâillement, du réflexe à la pathologie
Le bâillement : de l'éthologie à la médecine clinique
Le bâillement : phylogenèse, éthologie, nosogénie
 Le bâillement : un comportement universel
La parakinésie brachiale oscitante
Yawning: its cycle, its role
Warum gähnen wir ?
Fetal yawning assessed by 3D and 4D sonography
Le bâillement foetal
Le bâillement, du réflexe à la pathologie
Le bâillement : de l'éthologie à la médecine clinique
Le bâillement : phylogenèse, éthologie, nosogénie
 Le bâillement : un comportement universel
La parakinésie brachiale oscitante
Yawning: its cycle, its role
Warum gähnen wir ?
Fetal yawning assessed by 3D and 4D sonography
Le bâillement foetal

mystery of yawning 

haut de page




























haut de page





















haut de page




mise à jour du
6 novembre 2003
Dev Psychobiol
Yawning and Behavioral States in Premature Infants
Fiorenza Giganti, Piero Salzarulo
Psychology Department University of Florence Florence, ltaly
Marie J. Hayes Psychology Department University of Maine
Manjapra R. Akilesh Neonatology Department Eastern Maine Medical Center Sangor,
Télécharger l'intégralité de ce texte au format PDF
Development of behaviors in preterm infants: relation to sleeping and waking



-Giganti F, Hayes MJ, Akilesh MR, Salzarulo P. Yawning and behavioral states in premature infants. Dev Psychobiol. 2002;41(3):289-96.
-Giganti F, Hayes MJ Cioni G, Salzarulo P Yawning frequency and distribution in preterm and near term infants assessed throughout 24-h recordings Infant Behav & Development 2007;30(4):641-647
-Giganti F, Ziello ME Contagious and spontaneous yawning in autistic and typically developing children CPL 2009
-Giganti F, Zilli I. The daily time course of contagious and spontaneous yawning among humans. J Ethol 2011;29(2):215-216
-Giganti F, Zilli I, Aboudan S, Salzarulo P. Sleep, sleepiness and yawning. Front Neurol Neurosci. 2010;28:42-6.
-Giganti F, Salzarulo P. Yawning throughout life. Front Neurol Neurosci. 2010;28:26-31
-Giganti F, Toselli M, Ramat S. Developmental trends in a social behaviour: contagious yawning in the elderly. Giornale di Psicologia dello Sviluppo. 2012;101:111-117
Giganti F, Guidi S, Ramat S, Zilli I, Raglione LM, Sorbi S, Salzarulo P. Yawning: A behavioural marker of sleepiness in de novo PD patients. Parkinsonism Relat Disord 2013
-Zilli I, Giganti F, Salzarulo P. Yawning in morning and evening types. Physiol Behav 2007;91(2-3):218-222
-Zilli I, Giganti F, Uga V. Yawning and subjective sleepiness in the ederly. J Sleep Res 2008;17:3003-308
Tous les articles consacrés au bâillement foetal
Fetal yawning: all publications
Human, adult observational studies have established that yawning is a stereotyped action pattern that occurs at low frequency throughout the waking period (Provine, 1986a, 1986b). It is a complex behavior that is characterized by gaping of the mouth, a long inspiration followed by a shorter expiration. Like other stereotyped patterns, when initiated, it proceeds to completion with minimal influence from sensory feedback. Yawning often is accompanied by stretching, chewing, and suppression of locomotor activity. The frequent inclusion of whole body stretching and pende erection in animal studies has led te, the characterization of the behavior pattern as the "stretching-yawning syndrome" (Argiolas & Melis, 1998).

Yawning is facilitated by DA, acetylcholine, serotonin, nitric oxide, and adrenocorticotropic peptides which have been shown to stimulate oxytocinergic, paraventricular nucleus neurons and projections to hippocampus, pons, and medulla. Central nervous system (CNS) injection of adrenocorticotropic hormone (ACTH) or alpha-melanocyte stimulating hormone and related peptides increases yawning which continues for several hours; induced yawning is suppressed by opioid peptides and gamma aminobutyric acid (Argiolas & Melis, 1998; Ferrari, Gessa, & Vargiu, 1963; Gessa, Pisano, Vargiu, Crabai, & Ferrari, 1967).

In human adults, yawning bas been related to a change in activity or arousal state and is, therefore, tied to CNS arousal modulation. Yawning increases in the hour before bedtime and after waking (Provine, Hamernik, & Curchack, 1987c) and when subjects are watching or participating in uneventful, repetitious activities (Provine & Hamernik, 1986a). In a study of cab drivers working long shifts, yawning increased along with head, neck, and leg movements toward the end of the shift in the early morning hours (Sakai & Takahashi, 1975). Cortical electroencephalographic activity (a marker of increased arousal) increases during yawning (Bertolini & Gessa, 1981). These findings have been interpreted as evidence that yawning may be potentiated when an increase in arousal is needed to maintain alertness and vigilance in a waking state.

Yawning is a precocious behavior that is observed sporadically in fetuses as early as 12 to 14 weeks of gestational age (De Vries, Visser, & Prechtl, 1985). No changes in the incidence of yawns between 20 and 36 weeks of gestational age have been observed in the fetus by Roodenburg, Wladimiroff, van Es, and Precht (1991). In full-term infants, yawns were reported on the first day of life (Gesell, 1928). However, yawning in neonates is not reported in studies of spontaneous behavior in neonates, perhaps, in part, due to its exceptionally low incidence (e.g., Korner, 1969, 1973; Korner & Beason, 1972). In the only study to describe neonatal yawning in some detail, Wolff (1987) observed that infant yawns occur mainly near, or possibly anticipating, the onset of sleep, together with reddening of the eye sclera and drooping of the eyelids. State Il sleep follows yawning with a mean latency of 7 min.

The present study examined yawning during very early postnatal development in relation to contextual behaviors and spontaneous alterations in behavioral state. Since yawning is a precocious behavior during early fetal life, appearing many weeks before discemable sleep-wake states, its expression gradually may become linked to state. Further, since adult yawning is increased during sleep-wake transitions, we aimed to investigate if yawning may be a corollary of state transition and/or temporally close to sleep state during early development. We also examined the temporal association of yawning and the frequent bursts of generalized activity expressed by the neonate particularly during active sleep (AS) (Hayes, Plante, Fielding, Kumar, & Delivoria-Papadopoulos, 1994; Robertson, 1987). [...]

fetal yawn 

Yawning Incidence : Spontaneous yawning is rare. The rate of yawning across all infants in the 5-hr nighttime recording period averaged approximately one yawn/hr (median = 4.5 yawns/night). Table 3 shows that, using the quotient statistic described earlier, there were significant differences in the probabilities of yawning across the behavioral states, Friedman's X 2 (3) = 8. 0 1, p = .04. The highest prevalence of yawning was in the D state. Using the Wilcoxon test to compare all possible pairings of behavioral states revealed that a significantly lower quotient of yawning is present in QS coinpared to W (p = .04), D (p = .04), and AS (p = .04), No significant differences were found for the other quotient comparisons. Only 1 infant, the youngest (30 weeks of PCA), showed any yawning in QS.

Table 2. Coding Criteria

State Description

Active sleep (sommeil paradoxal) The infant's eyes are closed with periodic REMs occurring. Occasionally, eyes will be open during REMs. Motor activity occurs in bursts of 10 to 60 s (generalized whole body movements including facial, mouth, and head movements), with muscle tone remaining low between movement bures.Typically, respiration is irregular and costal, which is evident by observing the thorax and chest.

Quiet sleep(sommeil profond) The infant's eyes are shut. Motor activity is low, and muscle tone is moderate. Occasional mouthing or sucking movements may occur. The infant's respiration is even and abdominal in nature.

Wake (réveil) The infant's eyes are open, focused (i.e., but not rapid REM movements, often accompanied by head and arm movements; whole sequence resembles an orientation response), and scanning the environment. State may include crying or fussing. Motor activity may be high or low. Respiration is regular and tracks motor activity.

Drowse (endormissement) The infant's eyes open and close slowly. When open, the eyes are unfocused with a "heavy-lidded" appearance. Characteristics of both sleep and wake states are present, and respiration is even.

Yawn State : The distribution of sleep states for the infants in this study was typical for premature infants of this age (Giganti et al., 2001-, Holditch-Davis & Thoman, 1990; Meyers et al., 1998; AS: mean=48.9%, SD = 13.2; QS: mean = 26.4l'/c, SD = 8.9; D: mean = 11. 1 %, SD = 6.4; W: mean = 12%, SD = 6.6). Nursing interventions occurred on average for 4.8% (SD = 5.3) of the session.

To evaluate the temporal organization of yawning and state, yawns in each state were examined in relation to the preceding or following state (Table 4). QS was not included because there was only one example of a QS yawn in the dataset. Each yawn state (e.g., AS, D, W with at least one yawn) was examined for the state of the preceding or following 3-min epoch. The relative probability of the preceding or following state was examined for its relation to state stability or state change.

Preceding State : The preceding state for each yawn state was compared To nonyawn states from the same subjects. A two-way cross-tabulation statistic was used to examine the expected frequencies of each preceding state given each yawn state. Preceding state probabilities were found to be nonrandom and predicted by the yawn state, X2 = 85.7, df = 12, p <.000 1. Follow-up comparisons found that yawn states AS, X2 = 16.2, df = 2, p<.0001, or W, X = 11.6, df=2, p<.0001, were generally preceded by the same state. That is, preceding state and yawn state were stable. As shown in Figure 1, the preceding state to D yawn state was not stable, X2 = 2.6, df = 2, p <.27. That is, for D yawn state, the preceding state was either D (0.55) or W (0.40).

Next, AS, W, and D nonyawn states were examined similarly. These data were drawn randomly from the same sleep recording for each subject as the yawn-state analysis. As was found for the yawn states, nonyawn states predicted preceding state, X2 = 98.2, df = 12, p <.000 1. AS and W nonyawns states were stable in the preceding 3-min epoch, AS: X2 = 19.6, df=2, p<.0001; W: 100% of all preceding states were W. Unlike yawn D state, nonyawn D state was stable, i.e., usually preceded by D state, X 2 = 6.25, df = 2, p <.04.

Following State : Following states for yawn and nonyawn states were examined in the same manner. Overall, yawn state predicted following state, X2 = 108, df = 12, p < .0001. Follow-up comparisons revealed that yawn states predicted following states for all three states, AS: X 2 = 28.9, df = 2, p <.00 1; W: X2 = 7.36, df = 2, 2 p <.007; D: X = 6.4, df = 2, p <.O 1. These findings indicate that, unlike preceding states, the following states of yawn D state are not characterized by state change.

Nonyawn states also predicted following states, X2 = 103, df= 12, p <.000 1. Follow-up comparisons revealed that nonyawn states predicted following states for AS, X 2 = 9.8, df = 1, p >.001, and W (100% of the cases), but nonyawn D states were not stable, X2 = 3.25, df = 2, p <.20. As shown in Figure 2, the probabilities for following states in nonyawn D state are D=0.70, AS=0.20, and W=0.10.

Contextual Behaviors : The type of behaviors that were coded are defined in Table 5. These categories have been used previously to categorize movement type in premature infants under similar conditions (Hayes et al., 1994; Hayes, Smith, Herrick, Roberts, & Swanson, 1997) and are inspired by previous work by Smotherman and Robinson (1988).

In the analysis of contextual behaviors, the 30 s prior to and following each yawn were examined and compared to nonyawn periods. Nonyawn periods were matched to yawn periods for subject and state type. The results show that the 1 -min yawn and nonyawn periods matched for state were different in the total frequency of behaviors, yawn: X2 =17.33, df=4, p <.002; nonyawn: X = 8.00, df = 4, p <.09. Yawn periods were more behavior-rich than nonyawn periods. This difference was in the total frequency of behaviors only. It was not found when examining before and after periods separately or when the data were examined according to state type.

The only finding specific for behavioral typology was that facial movements were proportionally more common in AS yawn versus nonyawn 1-min contextual periods, X 2 = 7, df = 2, p <.03. D and W yawn states, as well as nonyawn AS, W, or D state, showed no différences in the proportion of separate behavior types.


In premature infants, the rare behavior of yawning is not randomly distributed with regard to behavioral state across the night. Thirty to 35 PCA infants are Ilemergent" with regard to state, Èe., rapidly developing the capacity to express reliable and stable state organization (Curzi-Dascalova, Peirano, & MorelKahn, 1988; Parmalee, Wenner, Akiyama, Schultz, & Stem, 1967). Nonetheless, yawns were observed in all states except QS. This result suggests that the necessary and sufficient conditions for the elicitation of yawns are not present, or may be actively inhibited, during QS.

As has been observed in adults, yawning prevalence in these infants showed a trend towards highest levels during D state (Provine, Tate, & Geldmacher, 1987b). One possibility suggested by earlier work is that yawning is seen during the D state and is associated with efforts to remain awake. In animal studies, yawning is associated with EEG activation and is triggered by peptides such as ACTH. In this study, our approach to examining the question of the effect of yawning on ongoing states of arousal was to examine yawning placement in the strearn of state organization.

We found that preceding state stability was temporally associated with yawn events in D state, but not in AS or W states. When compared to epochs of nonyawn D state, yawns in D state had a higher incidence of unstable preceding state. In yawn D state, the preceding state was either D or W. This finding suggests that when the preceding state is W state transitions to D may be associated with an increased probability of yawning.

Following states to a yawn state (whether D, W, or AS) were characterized by state stability. However, D state was unstable when yawns were not present. Under these conditions, the following state was either AS or W when state change occurred. It is proposed that the absence of yawning in D state may increase the probability of a state transition in the following 3-min epoch. The sequence analysis suggests that some yawning effects are state-specific. In D state, yawning is associated with recent state change (i.e., "preceding" state instability), but state stability in the following state epoch is sound for all yawn states. Interestingly, al! nonyawn states are stable for preceding state.

D state bas been viewed as a transitional state between wake and sleep (Salzarulo, 1973; Salzarulo, Giganti, Fagioli, & Ficca, in press). Yawning may serve as an arousal modulator in D that is potentiated by an unstable preceding state. Yawning under these conditions may increase arousal sufficiently to stabilize the D state, which is consistent with the position that nonyawn (but not yawn) D epochs were unstable for the sequence of following states. In total, these findings suggest that D state is stabilized by yawn events.

The temporal analysis of yawning revealed that behavior rate was increased in the 1-min window of yawns compared to nonyawn periods. This result is reminiscent of the finding that cab drivers yawned more and were more behaviorally active at the end of their shift. Both findings suggest that yawning is associated with an increase in nonspecific motoric activation.

Premature infants in this study were in AS between 50 and 70% of the time in the 5-hr recording period. W and D contextual behaviors were similar in type to AS periods, although lower in both frequency and variety. AS motor bursts are characterized by a rich variety of behaviors: stretching, facial movements, general movements, and so on. The temporal window of yawning in AS had more facial movements than AS periods without a yawn. This result suggests a spreading activation of facial motor patterning, as has been suggested during movement analyses in our earlier work (Hayes et al., 1994).

 Gianluca Ficca & Piero Salzarulo "Lo Sbadiglio Dello Struzzo" , Bollati Boringhieri, Torino, 2002