Forced expiratory maneuvers in spirometric or pneumotachographic assessments do only provide indirect and discrete measures of the airway caliber. They may also enhance bronchoconstriction by irritation of the airways or dilate the airways through deep inhalation, and they require attention and voluntary effort and are thus subject to multiple psychological influences (Smith et al., 1970; Philipp et al., 1972; Spector et al., 1976; Harm et al., 1984). In contrast, respiratory resistance monitoring with forced oscillation or the interrupter technique provides direct and more sensitive assessment of airway caliber and allows for a continuous monitoring of airway status throughout an experimental protocol, which is critical in most studies of emotion (Ritz et al., 2002). Because it relies on tidal breathing it is independent of effort or motivation and does not alter airway tone, and it can thus be expected to produce more valid results. In addition, the multiple-frequency forced oscillation technique also allows a separate estimation of large (central) vs. small (peripheral) airway effects (Smith et al., 2005).

Validity problems may also arise from variations in the timing of measurements which are sometimes inherent in the measurement technique. Discrete spirometric assessments can only be implemented before and after the task as they require attention and effort that would disrupt most tasks. Forced oscillation or interrupter measurements can be implemented for continuous measurements throughout tasks. Measurements need to be restricted to before and after tasks when the character of the task (e.g., involving speech production, use of hands) does not allow for continuous assessments. Two of three film studies that showed no significant change or low effect sizes have relied on pre-post assessments only (Table 2). An additional number of studies that showed weak or no effects have also relied on discrete assessments (Miklich et al., 1973; Rietveld et al., 1999; Kang and Fox, 2000; McQuaid et al., 2000; Put et al., 2004; Aboussafy et al., 2006; von Leupoldt et al., 2006). However, emotions have been conceptualized as phasic, transient events of a short duration (Levenson, 1988). Especially in the artificial laboratory setting an emotional state cannot be expected to continue after task-offset—if anything at all, recovery from the emotional state will be measured at this point in time. Indeed, studies that implemented recovery measurements with sensitive techniques have shown that in the first minute following the off-set of stimulation the observed airway constriction is only a weak reflection of that seen during the task (Kotses et al., 1987b; Ritz et al., 2000, 2011c). Taking into consideration that the initiation of a discrete assessment of lung function is time consuming and distracting itself (experimenter instructions, movements to reach for and apply mouth pieces and nose clips, adaptation to the new mode of breathing), such measurements in all likelihood are too far removed from the psychological phenomenon of interest, the resistance change elicited by an emotional state. Discrete measurement protocols operate under the assumption that the clinically relevant state is a tonic constriction of the airways that continues far beyond the actual activation of emotion, an assumption that may not necessarily be valid. The idea may not be ecologically valid with respect to actual dynamics of emotional experience in real life, where periods of emotional turmoil linked to the experience of asthmatic airway constriction may not consist of brief arousal of one emotional state, but of continuous transitions from one negative affective state to the next, or of repeated elicitations of such states in very close temporal proximity.

Given that autonomic outflow is not uniform across emotional states and has been shown to vary along dimensions of affect (Lang et al., 1993) or discrete emotional states (for review, see Kreibig, 2010), it would be simplistic to expect that many different psychological states lead to the same outcome in terms of responding of the airways. Nevertheless, induction of a variety of emotional qualities with standardized induction techniques has been shown to produce constriction of the airways.

For unpleasant affective states, constriction is most uniform across film- and picture induction techniques, which have sometimes been conceptualized as passive coping tasks (Lehrer et al., 1996), because the subject passively endures an aversive stimulus. In contrast, stress tasks often require a more active mode of operation from the subject, requiring the mobilization of information processing capacities, speech, posture changes, or smaller manual activities. Such active action orientation could already mobilize some of the autonomic changes typically seen for these activities, as demonstrated by studies on exercise imagery (e.g., Calabrese et al., 2004). They may also constitute active coping challenges (Obrist, 1976) with strong components of sympathetic arousal that could counteract bronchoconstriction. It also needs to be considered that the emotional quality of such tasks can be mixed or ambivalent and can change with perceived coping success. Classical stress tasks are therefore not comparable with tasks that elicit discrete emotional states more purely, such as carefully pre-evaluated movie scenes (e.g., Gross and Levenson, 1995).

In addition, intensity of the elicited state can also determine airway constriction. Illustrated with a between-individual example, surgery films elicit massive differences in anxiety and disgust experience between individuals with blood-injection injury phobia and asthma patients or healthy controls (Ritz et al., 2011c). Consequently, we observed the strongest respiratory resistance increase in blood-injection-injury phobia individuals (25–30%) compared to healthy controls (11–17%), with asthma patients (21–23%) taking an intermediate position.

It is also notable that hypnotic suggestion tasks often used in earlier studies (e.g., Heimlich et al., 1966; Tal and Miklich, 1976; Ben-Zvi et al., 1982) have elicited considerable bronchoconstriction or bronchodilation (typically 15–20% FEV₁ change on average) with induction of negative or positive emotional states, respectively, despite mostly using discrete measures of spirometry. It is possible that the emotional states elicited by a hypnotic technique were more intense and in most studies had continued throughout lung function assessments. On the other hand, another study of Clarke (1970) observed FEV₁ decreases >20% in two of three patients with hypnotic induction of anger and fear only when combined with suggestions of an asthma attack. However, in these studies self-report of the emotional experience was often not measured or not reported in a way that would allow systematic comparison with more recent emotion-induction studies. It is also possible that less sophisticated medication regimens of patients in earlier studies allowed for stronger airway reactivity. A complication of the hypnotic induction is that there are substantial individual differences in susceptibility to suggestion (Leigh et al., 2003), which may make emotion-induction success more variable compared to more recent techniques using films or imagery.

Finally, it needs to be considered that the character of the employed stress challenges has been acute, as compared to more persistent, repeated, or chronic states of stress that are for humans more likely encountered in daily life, rather than in a controlled laboratory setting. Both studies with animals (e.g., Forsythe et al., 2004; Kang and Weaver, 2010) and humans (Kullowatz et al., 2008; Marin et al., 2009) have shown differential responding in airways and asthmatic pathophysiology to states of acute vs. chronic stress. Unfortunately, there is still little consensus in the literature on the exact definition of acute vs. chronic stress challenges.

From a psychophysiological standpoint, some of the laboratory research on emotion or stress induction and airway responding has been surprisingly limited. Even though, most studies implemented state-of-the-art physiological assessments, the psychological assessment side has been unsophisticated or completely lacking. Thus, only about half of the stress-induction experiments had some type of self-report instruments in place to capture the actual emotional or stress experience of the subjects. Under this scenario, experiential state of the subject is simply being inferred from the potential of the laboratory task to elicit that state, which is reminiscent of an early behaviorist black-box view of higher mental processes (e.g., Obrist, 1976). However, uniform responding of participants to experimental emotion-induction scenarios cannot be expected (Lang, 1994). In addition, given that most modern scientific theories define emotion as a state that unfolds on multiple levels of observation, it appears mandatory to include at least experiential and physiological levels for a minimal description of the phenomenon. Emotion theory would also hold that none of these levels of description can serve as a validation of the other or as ultimate indicator for the presence or absence of emotion (Levenson, 1988). At the minimum, the inclusion of the experiential level would serve as an additional indicator of the validity of the various experimental tasks used in this research and help elucidating their emotional quality. Sole reliance of physiological indices, such as, blood pressure or skin conductance increases, is not sufficient for identifying individuals that respond emotionally to a laboratory stressor (e.g., Horton et al., 1978; McQuaid et al., 2000; Laube et al., 2003).

Perhaps the best evidence to date comes from laboratory studies that have implemented pharmacological blockade. Given that the cholinergic pathway is the main bronchoconstrictor (Canning and Fischer, 2001), efforts have been concentrated on blocking muscarinic impulse transmission at the airway smooth muscle. Already in 1969, McFadden et al. demonstrated that blockade with intravenous atropine abolished airway resistance increases to bronchoconstrictive suggestion. Similarly, Neild and Cameron (1985) blocked bronchoconstrictive suggestion effects in a selected group of patients that had previously shown a clinically significant fall in FEV₁ to this type of suggestion. This may indicate central vagal excitation or altered sensitivity to normal vagal outflow at the end-organ level. In line with the latter interpretation, Horton et al. (1978) showed a significant association of suggestion effects with hyperreactivity of the airways tested by both methacholine and histamine. Thus, centrally mediated vagal excitation can only be viewed as a major pathway of bronchoconstrictive suggestion effects if airway hyperreactivity to these pharmacological challenges itself is also viewed as at least partially mediated by central vagal pathways (e.g., Haxhiu et al., 2005).

We recently examined airway responses to emotional films and pictures under both placebo and ipratropium inhalation in asthma patients and healthy controls (Ritz et al., 2010b). Increases in resistance were strongest under negative emotional stimulation in patients and were significantly attenuated by cholinergic blockade. Because blockade by ipratropium does not affect cholinergic transmission in the glottis (Widdicombe, 1986), substantial upper airway effects can therefore be excluded as mechanisms of resistance changes. Resistance increases were also not significantly associated with airway reactivity to methacholine, dilatory response to ipratropium at baseline, or basal airway inflammation measured by exhaled nitric oxide. The finding is consistent with that of Lehrer et al. (1996) who did not observe an association between responding to mental arithmetics, reaction time task, or aversive and surgery film viewing with responsiveness to methacholine in asthma patients or healthy controls. Further correlational findings from two studies that have used stimulation with affective pictures and self-referring statement suggest a role of vagal outflow in at least some instances of emotion elicitation. Under conditions of depression induction, respiratory resistance was positively correlated with respiration-corrected respiratory sinus arrhythmia (as index of cardiac vagal activation) in asthma patients (Ritz et al., 2001, 2005). Thus, central vagal excitation appears to be the critical final pathway linking psychological processes to airway responses to viewing of emotional film and picture stimuli.

Modulation of airway tone by sympathetic effects is important, as demonstrated readily by β-adrenergic bronchodilator medication. Action of circulating catecholamines on β₂-receptors appears to be the primary source of endogenous sympathetic bronchodilation in humans because of a lack of direct sympathetic innervation of the airway smooth muscles (Barnes, 1986; Canning, 2006). However, additional modulation of cholinergic outflow by sympathetic nerves may take place at the level of the ganglia. To date, only sporadic correlational findings are available that may suggest an association of airway responses to emotions or stress with sympathetic activation. Skin conductance, a measure associated with sympathetic cholinergic regulation of sweat gland activity, has shown similar patterns of modulation across emotional states as respiratory resistance (Ritz et al., 2000) and has produced significant positive correlations with respiratory resistance increases between individuals (Butler and Steptoe, 1986; Ritz et al., 2011c). The positive association between an index of sympathetic excitation and respiratory resistance defies simple interpretations and it is unclear whether it signifies an instance of autonomic nervous system fractionation (Berntson et al., 1991), is a sign of a more widespread damage to cholinergic neurotransmission by allergic processes (Kaliner, 1976; Lehrer et al., 1996), or only constitutes a correlation of processes that are functionally unrelated but associated through their relationship with an unknown third variable. Effects of sympathetic excitation may be more prominent in experimental or real-life stressors compared to emotion-induction tasks. Kang and Fox (2000) observed that post-stressor PEF was positively associated with plasma epinephrine levels.

Other autonomic pathways, such as, the nonadrenergic-noncholinergic system have not been studied systematically in humans in the context of stress induction. A role for tachykinins (such as substance P) secreted from sensory nerve endings has been suggested in mice (Joachim et al., 2003, 2004). Sensitized and allergen challenged mice were found to show stronger airway hyperreactivity and eosinophil infiltration following 24 h of sound stress compared to no stress condition, a response that was eliminated with of a neurokinin-1 receptor antagonist. As a hypothetical pathway, stress could enhance activity of the hypothalamic-pituitary- adrenal axis, which has been shown to raise nerve growth factor levels, and that in turn may induce substance P secretion from sensory nerve endings in the airway epithelium. In a cross-sectional study with humans, serum level of brain derived neurotrophic factor has been shown to be elevated in asthma with self-reported chronic stress (Joachim et al., 2008).

A general problem with studying autonomic pathways is a considerable amount of target organ specificity, which has been demonstrated for both parasympathetic and sympathetic systems (Morrison, 2001; Ritz, 2009). Thus, indices of autonomic function derived from other organ sites (e.g., cardiovascular system or sweat glands) cannot necessarily be expected to inform about autonomic regulation of the airways, although target specificity may vary situationally and across states of health and disease.

Given that the airway smooth muscles express receptors for a variety of hormones (Thomson et al., 1996), multiple pathways of influence are conceivable, but research in the context of stress and emotion effects on the airways is still in its infancy. In addition to catecholamines, other endocrine parameters have only been explored sporadically in prior studies. Cortisol levels following an acute laboratory free speech and mental arithmetic stressor were negatively associated with PEF levels in one study with asthmatic and non-asthmatic students (Kang and Fox, 2000), but no association of spirometric lung function changes with cortisol changes from low stress to academic stress periods were observed in atopic or non-atopic students (Höglund et al., 2006).

Ventilatory changes are known to influence airway smooth muscle tone in a number of ways. Deep inspiration is a potent bronchodilator (Nadel and Tierney, 1961; Krishnan et al., 2008), but this effect may be absent in more severe asthma (Scichilone et al., 2007). Airway resistance is inversely related to lung volume and elevations in functional residual capacity are associated with bronchodilation (Forster et al., 1986). Increases in flow can be coupled with increases in airway resistance (Hida et al., 1984), although this association may depend on the behavioral context, with increases in flow related to skeletal muscle activation showing the opposite effect (Baker and Don, 1988). Hypocapnia is also known to constrict the airways through pathways involving autonomic vagal excitation and local airway effects (Newhouse et al., 1964; O'Cain et al., 1979) and this effect is more pronounced in asthma than in healthy controls (van den Elshout et al., 1991). Hyperpnea (or isocapnic hyperventilation) constricts the airways through drying and/or cooling of the airways, which is viewed as the major pathway of exercise-induced bronchoconstriction (McFadden and Gilbert, 1994; Anderson, 2010).

Given that ventilatory influences have often been suspected as a major pathway in emotion-induced airway responding (Clarke and Gibson, 1980; Knapp, 1989), surprisingly few studies have examined such influences. In suggestion studies, only sporadically measurements of timing and volume parameters of the respiratory cycle were included, with two studies showing no changes (Weiss et al., 1970; Butler and Steptoe, 1986) and one showing decreases in tidal volume (V_T) following bronchoconstrictive suggestion (Put et al., 2004). In emotion-induction studies, ventilatory changes have not been found to be consistently related to changes in respiratory resistance. In one study using unpleasant and surgery-related film material for stimulation, V_T changes showed a systematic negative association with resistance, with a 100 ml increase in V_T leading to a 0.005 kPa·L⁻¹·s decrease in oscillatory resistance, but overall levels of resistance and emotion-induced changes of resistance throughout film clips were not affected by V_T change (Ritz et al., 2012).

Ventilatory influences could be most prominent in states involving marked emotional expressions, such as crying (Weinstein, 1984) and laughing (Liangas et al., 2004). Similarly, bronchoconstriction following amusement park rides (Rietveld and Van Beest, 2006) could be related to such factors (screaming, laughing). In addition to airway drying and/or cooling, stimulation of irritant receptors through high flow rates could be another mechanism responsible for such airway constriction in asthma (Hida et al., 1984). However, studies of emotional expression effects have typically not included detailed assessments of the respiratory pattern or gas exchange. Evidence on stress-induced hypocapnia as a bronchoconstrictor has mainly come from early case studies (Herxheimer, 1946). One earlier study observed strong increases in minute ventilation in patients imagining asthma-relevant scenes, but no measures of gas exchange were included (Clarke and Gibson, 1980). In a recent study using a psychosocial laboratory paradigm (free speech and mental arithmetic task under evaluative pressure), no evidence for hypocapnia was found in individuals with asthma or healthy controls during quiet speech preparation and after termination of the task (Ritz et al., 2011b).

Animal studies suggest that the airway-related vagal preganglionic neurons in the rostral parts of the nucleus ambiguus and the rostral dorsal motor nucleus constitute the central integrators of multiple CNS influences on airway smooth muscle tone (for review, see Haxhiu et al., 2005). Input is provided from higher CNS centers in forebrain, midbrain and hindbrain, including prefrontal cortex, central nucleus of the amygdala, hypothalamus, and periaqueductal gray (PAG), pons, and nucleus tractus solitarii. A central role seems to be attributed to the PAG, which is also known for its coordination of active and passive coping responses to behavioral challenge. Activation of its ventrolateral portion leads to airway smooth muscle relaxation (Haxhiu et al., 2002) through a GABAergic inhibitory pathway to the vagal preganglionic neurons or through the parabrachial nucleus (Motekaitis et al., 1994, 1996). Another pathway seems to exist from the paraventricular nucleus of the hypothalamus, which has neurons that project directly to the airway-related vagal preganglionic neurons and that augments their activity through release of oxytocin and arginine vasopressin (Kc et al., 2006). Projections from the amygdala to the paraventricular nucleus could then constitute a pathway for inducing bronchoconstriction by negative emotions.

Although there is some exploration of pathways in animal models, little is known about relevant brain regions in humans. Most research stems from studies of respiratory sensation using stimulation by added resistive loads (e.g., Peiffer et al., 2001; von Leupoldt et al., 2009). However, because there is typically only a low to moderate association of respiratory sensations with actual airway constriction and large interindividual differences in this association (e.g., Apter et al., 1997; Ritz et al., 2010c), the value of these studies for informing about CNS pathways of airway constriction is limited. In this context, activation of the PAG has been linked to reduced dyspnea perception (von Leupoldt et al., 2009) and increased gray matter volume of this structure has been associated with longer duration of the asthmatic disease (von Leupoldt et al., 2011). In addition, stimulation of the PAG and the subthalamic nucleus with deep-brain electrodes in patients with movement or pain disorders showed improvements in some aspects of spirometric lung function (14% increase in PEF, but no substantial change in FEV₁ or Functional Vital Capacity; Hyam et al., 2012). Taken together with findings from animal research, PAG functioning may have a central role in mediating emotion-induced airway responses. Exploring brain regions linked to emotion-induced airway constriction will also require careful consideration of neural pathways involved in ventilatory responses to emotion (Evans, 2010), which will most likely show a certain degree of overlap.

Environmental exposure and disease states of the airways may impact vagal preganglionic neurons or pathways that connect with them and may thus alter CNS processing of various stimuli including those of a psychological character. Prolonged exposure to allergens or noxious agents could damage the GABAergic inhibitory pathway that provides outflow from the PAG (Haxhiu et al., 2002). In a murine model of asthma, evidence was found for a downregulation of noradrenergic inhibitory traffic to the preganglionic vagal nuclei following sensitization and allergen exposure (Wilson et al., 2007). Particulate matter exposure has been shown to reduce excitability of cardiac vagal motor neurons in the nucleus ambiguus of mice (Pham et al., 2009). Exposure to allergens, ozone, or second-hand cigarette smoke in various animal models have been related to plasticity of the nucleus tractus solitarius, which coordinates respiratory motor output of a number of defensive reflexes through vagal pregangionic neurons that may be exaggerated in consequence and thus lead to stronger bronchoconstriction or cough (Chen et al., 2001, 2003; Bonham et al., 2006; Sekizawa et al., 2011). An elevated tendency to constrict the airway in situations that require coping with affective challenges may be the consequence. To date, studies linking CNS pathways with airway responses to emotional stimuli are missing, also because of technical challenges of measuring mechanical lung function in a neuroimaging environment.

Inflammatory processes in the airways related to the late-phase response of allergen challenge, such as increase in sputum eosinophils and reduced glucocorticoid response, have also been shown to be selectively associated with responses of the anterior cingulate cortex and insula to asthma-related compared to neutral words (tested relative to a non-inflammatory methacholine challenge) (Rosenkranz et al., 2005), possibly reflecting the heightened awareness of the inflammatory condition. Bronchoconstriction during the late phase response was also positively associated activation of the insula, which is known for its role in visceral perception.

Recent human studies have highlighted the role of stress and emotion in modulating systemic markers of inflammation in asthma (e.g., Kang et al., 1997; Marin et al., 2009). The dominant paradigm of this research has been the stimulation of white blood cells to measure cytokine activity in samples taken during stress vs. non-stress conditions as well as asthmatic and healthy subjects, thus essentially informing about the capacity of stress to modulate allergic responses. Following a similar paradigm but studying localized effects in the airways, Liu et al. (2002) demonstrated that late-phase eosinophil levels (measured from induced sputum) to allergen challenge are exaggerated by academic test stress and that such responses are associated with greater bronchoconstriction. On the other hand, purely observational research not involving allergen challenges found that in individuals with asthma, acute negative affect per se was associated with a reduction in FEV₁, whereas increases in daily hassles across a 3-month period were associated with improvements in FEV₁ (Kullowatz et al., 2008). These associations were mediated by increases and decreases, respectively, in FeNO as an indirect measure of airway inflammatory status. In this study, findings were only observed with FEV₁, whereas no association was observed with respiratory resistance measured by impulse oscillometry. This may suggest that experimental emotional stimuli such as brief film clips, which typically show robust but transient increases of respiratory resistance, capture an aspect of emotion-induced airway responding that is different from the constriction observed due to current negative affect, which may also be accompanied by inflammatory processes. The lack of association of film-induced resistance increases with FeNO (Ritz et al., 2010b) is consistent with this interpretation. It is possible that negative emotion-induced airway constriction is based more on phasic change in autonomic outflow, whereas current negative affect as a mood state leads to more tonic change in inflammatory parameters that may also affect the airways and/or only the vigor with which the subject performs the forced expiratory maneuver.

The fast onset of the phasic component of airway constriction to emotional provocation by film stimuli would also preclude a release of bronchoconstrictive mediators from mast cells as underlying mechanism. Corticotropin-releasing hormone secreted from postganglionic sympathetic nerves is known to degranulate peripheral mast cells (Pang et al., 1998; Cao et al., 2005) leading to the secretion of vascular endothelial growth factor (VEGF), which in turn has been linked to bronchoconstriction (Kanazawa et al., 2002). However, airway constriction related to allergen challenge typically develops over minutes rather than seconds (Grainge et al., 2011). These pathways could therefore be more likely involved in tonic airway responses to experimental or real-life stressors. VEGF has been shown to be elevated with negative affect (e.g., Asberg et al., 2009). VEGF from exhaled breath condensate was also found to be increased during academic exam stress in both healthy and allergic rhinitis (including asthmatic) individuals, but only in the latter group were increases in negative affect positively associated with VEGF increases (Trueba et al., 2012). Thus, the role of early-phase response inflammatory mediators secreted from mast cells, including histamine, leukotrienes, and prostaglandins, deserves further scrutiny as a pathway affecting airway tone in acute stress. Additionally, the role of airway nitric oxide, which has been shown to be elevated in response to acute stress and negative affect (Kullowatz et al., 2008; Ritz et al., 2011a), requires further study in this context. Nitric oxide resulting from endothelial and inducible nitric oxidase activity is thought to mediate a range of pulmonary effects of VEGF, including airway hyperreactivity (Bhandari et al., 2006). Nitric oxide has been shown to be associated with either bronchodilation or bronchoconstriction depending on its source (Ricciardolo et al., 2004).

In summary, at the present stage of research the two most plausible pathways linking emotion or stress with bronchoconstriction are autonomic vagal and ventilatory influences, in particular airway cooling and/or drying. It should be noted that these pathways would likely result in different temporal characteristics of bronchoconstriction (Figure 2). Airway constriction by vagal excitation would show a fast onset within the first 10–20 s of the beginning of stimulus presentation and subside with 1–2 min of stimulus off-set. Constriction due to emotion-induced hyperpnea would be more likely to follow a pattern similar to exercise or cold air challenges, which are typically administered for 6–8 min (Weiler et al., 2010), but the intensity of the constriction would depend on the level of ventilation the person reaches. So far empirical evidence has only been provided for the first type of response trajectory. Little is known about the dynamics and mechanisms of tonic components of airway constriction to emotion or stress and other temporal characteristics such as cumulative effects of repeated stimulation, sensitization, or habituation.

In order to study airway dynamics during changing experiential and behavioral conditions, constriction should be monitored throughout film presentations while the emotional process unfolds. Forced oscillation and interrupter techniques are well suited for this purpose (Ritz et al., 2002). Measurements following off-set of the stimulation material may be insensitive and conceptually problematic because they capture effects of recovery from stimulation rather than the effect of stimulation itself. Lung function assessments that require a forced expiratory maneuver are not suitable because they are discrete rather than continuous and interfere with the phenomenon of emotion.

Additional monitoring of ventilation is strongly recommended to explore potential ventilatory mediation of resistance increases, by e.g., reductions in V_T or hypocapnia.

Peak response is reached within the first 1–2 min of presentation, thus, film sequences as brief as that may suffice; however, trajectories of constriction across longer presentations (up to 5 min) have also been shown to provide useful information (Ritz et al., 2012).

As a reference, neutral film material is preferable, as it controls for non-specific effects of film viewing as such and helps to isolate experimentally the emotional aspect of the stimulation material. To control for within-session changes in tonic levels of airway resistance, it may be advantageous to implement intermediate baselines or brief pre-presentation measurements of 1–2 min duration.

Self-report of emotion should always be captured in addition to resistance monitoring, because it provides concurrent evidence for manipulation success. Video monitoring of facial expression (Ekman and Friesen, 1978) or electromyographic recordings of activity over facial muscle sites (Fridlund and Cacioppo, 1986) is an excellent additional level of behavioral evidence, although technically more demanding and sensitive to audience effects (imagined or real presence of individuals co-viewing the material with the participant; Fridlund, 1991; Hess et al., 1995).

Emotion is a construct with multiple facets that are not necessarily highly correlated. Rating of emotional states should therefore employ measures that reflect this plurality of states. The two dominant models of the structure of emotion either distinguish discrete emotional states (e.g., anger, sadness, and joy; Izard, 1991) or dimensions of emotion (e.g., valence and arousal; Russell, 1980; Lang et al., 1990). Using both approaches to measurement in the same study may be advantageous because co-variation of respiration and respiratory resistance has been observed with both models of emotion (Boiten, 1998; Ritz, 2004; Gomez et al., 2005). Respiratory sensations are another important domain of experience to be measured in studies of airway psychophysiology (Ritz et al., 2002). A solid body of literature has demonstrated multiple facets of respiratory sensations or dyspnea that require multidimensional assessment strategies (e.g., Kinsman et al., 1973; Simon et al., 1989; Lansing et al., 2009; Parshall et al., 2012).

Although there is evidence that a wide range of emotional states leads to airway constriction, film material of the blood-injection-injury type, or scenes of bloody surgery from medical education material, appears to be particularly potent (Ritz et al., 2011c, 2012). This also applies to individuals with asthma, who do not respond stronger to film scenes of asthma symptoms or attacks than to various other types of unpleasant film scenes.

Bronchodilatory medication should be discontinued prior to the assessment. At least 8 h have been recommended for short-term bronchodilators, 16 h for long-term bronchodilators, and 24 h for leukotriene modifiers (Ritz et al., 2002). Longer wash-out periods of up to 1 week will be needed for Tiotropium, a long-term anticholinergic agent that shows extremely slow dissociation from M₃ muscarinic receptors (Barnes et al., 1995).

Stable psychological characteristics, such as, previous experience of psychological asthma triggers (Ritz et al., 2006, 2010b, 2012), self-efficacy in asthma management (Campbell et al., 2006), or a tendency to use defensive coping strategies (Feldman et al., 2002) have been shown to moderate effect of emotional stimulation on the airways. Controlling for these and other psychological trait and temperament variables is recommended for future studies. Similarly, moderator effects of comorbid psychiatric disorders, particularly anxiety and mood disorders, on emotion-induced airway responses require further study, as two studies of comorbid asthma and panic disorder have demonstrated (Carr et al., 1996; Dorhofer and Sigmon, 2002).

The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.