INTERNATIONAL

JOURNALOF
PSYCHOPHYSIOLOGY
ELSEVIER International Journal of Psychophysiology 17 (1994) 103-128

Review article

Emotions and respiratory patterns: review and critical analysis

Frans A. Boiten I,*, Nice H. Frijda , Cornelis J.E. Wientjes b
aDepartment of Experimental Psychology, lJnic~er.sity of Amsterdam, Roetrrsstraat 15, 1018 WB Amsterdam, The Netherlands
h TN0 Human Factors Research Institute, Soesterberg The Nether1and.r

Received 4 January 1993: revised 5 January 1994; accepted 15 February 1994

Abstract

The literature on emotions and respiration is reviewed. After the early years of experimental psychology,
attention to their relationship has been sparse, presumably due to difficulties in adequate measurement of
respiration. The available data suggest nevertheless that respiration patterns reflect the general dimensions of
emotional response that are linked to response requirements of the emotional situations. It is suggested that the
major dimensions are those of calm-excitement, relaxation-tenseness, and active versus passive coping. Research on
the emotion-respiration relationships has been largely restricted to the correlates of respiration rate, amplitude, and
volume. Finer distinctions than those indicated may well be possible if a wider range of parameters, such as the form
of the respiratory cycle, is included in the investigation.

Key words: Emotions; Respiratory patterns; Critical analysis

1. Introduction tional states. Thus, evidence for links between

emotion and respiration may lead to the plausible
Emotions are linked to respiration. Panting in supposition that different emotions are associ-
fear, rage and sexual excitement, yawning during ated with different respiratory patterns. That in-
boredom, and sighing in melancholy, relief or deed has been a predominant hypothesis in ear-
distress are common occurrences. Everyday lier days of research in the psychophysiology of
speech contains numerous expressions which as- emotions (Feleky, 1914; Rehwoldt, 1911), and it
sociate breathing with emotional states (breath- has recently been advocated again (Santibafiez
taking, a sharp intake of breath, to be and Bloch, 1986; Bloch et al., 1991). The aim of
breathless with surprise, to catch ones breath, the present paper is to review the literature on
to breathe again freely), and terms like sigh- this subject, and to evaluate the current status of
ing, gasping, snorting and sobbing all this hypothesis.
have direct emotional connotations. In the era of Contemporary research on the relation be-
the silent movies, actors typically employed exces- tween emotions and respiration is sparse. The
sive breathing as a technique to express emo- respiratory response system has never aroused
the same degree of interest as, for instance, car-
diovascular variables and skin conductance. Char-
* Corresponding author. acteristically, Greenfield and Sternbachs Hand-

0167~8760/94/$07.00 0 1994 Elsevier Science B.V. All rights reserved

SSDI 0167-8760(94)00029-E
book of Psychophysiology (1972), that for so long upon breathing patterns in normal individuals. To
has been somewhat of a standard text, does not achieve this purpose we will initially consider the
contain a chapter on respiration. The main rea- nature of emotion-respiration relationships within
sons for the infrequent inclusion of respiratory the context of a number of emotional models.
measures in psychophysiological studies appear Since many important questions seem to be
to be primarily based on practical and method- awaiting adequate measurement methodologies
ological considerations (Wientjes, 1992, 1993). the next section of the paper is devoted to the
Traditional respiratory measurement techniques discussion of respiratory parameters that have
are often either too intrusive, too complicated, or been used in the analysis of breathing patterns,
too imprecise to provide adequate assessment of and the recording techniques that have been cm-
respiratory responses. Another reason for the ap- ployed in the measurement of respiration. We
parent lack of enthusiasm about respiratory mea- will then discuss the evidence regarding effects of
sures might be that the respiratory response is the induction of (negative and positive) emotional
not a unitary phenomenon, but varies in a com- states, and of cognitive factors, upon respiratory
plex manner along a number of time and volume parameters. Cognitive factors will be discussed in
dimensions (Stein and Luparello, 1967). Thus, so far as they may vary along with particular
measurement of respiratory responses requires emotional conditions, and may be held responsi-
techniques that provide sufficient precision with ble for differences found between such condi-
regard to the assessment of respiratory time and tions. Next, the available evidence regarding the
volume parameters, as well as relatively easy ap- differentiation of emotional states will be re-
plicability and unobtrusiveness. As we will see, viewed. Finally we will try to draw some conclu-
such techniques have only recently become avail- sions, in which we will come back to the other
able. models that were discussed. We have also in-
Links between emotions and respiration have cluded some of the older literature in this review.
been established in several distinctive research Although conducted with primitive techniques,
areas. First, there is evidence that emotions in and lacking adequate statistical treatment of the
general may influence respiration, and it might be data, these findings themselves are often interest-
that different emotional states give rise to differ- ing enough, and have generated suggestive hy-
ent breathing patterns. Second, breathing pat- potheses.
terns have been shown to differentiate between
diagnostic groups such as neurotics, psychotics
and normals (e.g., Clausen, 19.51; Coppen and 2. The nature of emotion-respiration relation-
Mezey, 1960; Damas-Mora et al., 1976, 1982; ships
Skarbek, 1970), and have served as a diagnostic
tool for assessing such groups. Third, some inves- With regard to the relationship between emo-
tigations have suggested that modifying and regu- tions and physiological parameters, three differ-
lating ones respiration may help to reduce sub- ent types of models can be distinguished. The
jective distress (e.g., Clark et al., 1985; Grossman specificity model hypothesizes that different emo-
et al., 1985; Holmes, McCaul and Solomon; 1978). tions are characterized by different physiological
Fourth, other studies have shown that respiration response patterns. These patterns differentiate
is related to personality or constitutional type, or one type of emotion from another. This hypothe-
general emotional orientations (Morgan, 1983; sis, obviously in line with the James-Lange theory
Shershaw et al., 1973; Shershaw et al., 1976; of emotion (e.g., James, 18841, has recently found
Wientjes, 1993). new supporters, who reported positive evidence
Although we will briefly mention a number of regarding autonomic response patterns differenti-
other studies, this review focuses primarily upon ating the six basic emotions of joy, sadness,
studies that deal with the first kind of relation- fear, anger, surprise, and disgust (Levenson, Ek-
ship, that is, with effects of (induced) emotions man and Friesen, 1990).
 F.A. Boiten et al. /International Jourrml of Psychophysiology I7 (1994) 103-128 105

The dimensional model assumes that physio- description or definition of these emotions
logical parameters or patterns correspond to gen- (Frijda, 1986). In actual practice, specific in-
eral dimensions of emotional feeling or emotional stances of particular emotions may be distributed
response that cut across emotion boundaries. widely over the space, even if typical instances, or
Relevant dimensions of feeling are pleasant-un- the majority, occupy a point or restricted region.
pleasant, active-passive, and calm-excited, the di- For instance, typical anger involves more strenu-
mensions proposed by Wundt (1902) and (at least ous effort or higher activation than typical sad-
the first two) supported by more recent dimen- ness or joy; yet, there exist instances of passive
sional analyses of mood (e.g., Russell, 1980; Wat- anger (e.g., anger in, Spielberger et al., 1985),
son and Tellegen, 1985). of frenzied sorrow, and of vigorous joy. Correla-
The third model, response requirement model, tions between physiological patterns and emo-
is not incompatible with the dimensional one. In tions, therefore, do not constitute decisive evi-
this model, physiological parameters or patterns dence that these patterns in a strict sense be-
are thought to correspond to particular response long to these emotions. They may be due to the
requirements of the given emotional situation. particular sampling of instances of the emotions
Response requirements underlie the functional concerned.
categories of flight and fight (Cannon, 19291, and Similar difficulties arise in distinguishing be-
more recently proposed functional dimensions tween a dimensional and a response requirement
like intake-rejection (Lacey and Lacey, 1970) and model. Theoretically, it makes a difference
passive versus active coping (Obrist, 1981). A whether a particular pattern of physiological re-
related formulation is that physiological patterns sponse is thought to reflect, say, pleasant feeling,
might correspond to appraisal patterns, or the or readiness to accept stimulation; the first points
patterns of stimulus evaluation checks performed to an expressive correlation, the second to a
on the eliciting situation (Frijda, 1986; Scherer, functional one, the relation of which to physiolog-
1986). Appraisal patterns and response require- ical parameters is more readily understood. How-
ments are closely related notions, since the for- ever, one may expect that the two types of dimen-
mer are seen as the individuals perceptual basis sion are correlated; in fact, explanations of physi-
for assessing the latter. The appraisal version of ological response patterns in terms of general
this model has been applied to generate hypothe- response dimensions like arousal, activation, and
ses about the psychological correlates of voice inhibition (Pribram, 1981) fall midway between
intonation patterns (Scherer, 19861, an area obvi- the two models. Only to the extent that the
ously related to that of respiration. In both ver- dimensions of feeling diverge empirically from
sions, physiological response patterns again are the response requirement or appraisal dimen-
thought to correspond to emotion dimensions sions (pleasant-unpleasant from, say, intake-re-
that cut across the boundaries of emotion cate- jection), can these classes of models be empiri-
gories. Active coping, or the appraisal of antici- cally separated. The available evidence on emo-
pated effort, for instance, might be common to tions-respiration relationships will be reviewed
certain forms of anger and certain forms of fear. primarily along the lines of the specificity model.
It will not be easy to distinguish between the
three types of models. Physiological response pat-
terns may be correlated with particular emotion
classes without corresponding to these emotions 3. Respiratory parameters
as such. One reason is that, while different emo-
tions are found to be consistently linked to partic- Respiratory parameters can be categorized in:
ular locations in the space formed by two or three (1) volume and timing parameters, (2) measures
of the dimensions indicated (Russell, 1980; Wat- regarding the morphology of the breathing curve,
son and Tellegen, 1985; Woodworth and Schlos- and (3) measures reflecting gas exchange (see
berg, 1955), these locations do not exhaust the Wientjes, 1992).
106 F.A. Boiten et al. /lntemational Journal of Psychophysiology 17 (1994) 103-128

3.1. Volume and timing parameters drive mechanism, and Ti/Ttot reflects the peri-
odicity of the timing mechanism (Gautier, 1980;
The respiratory cycle consists of an inspiration Milic-Emili and Grunstein, 1976; Milic-Emili et
phase, followed by an expiration phase. There al., 1981). This latter measure is often referred to
sometimes is a pause between termination of the as inspiratory duty cycle (a term that is used in
inspiration phase and onset of expiration, and electrical engineering to denote the fraction of a
often between termination of expiration phase periodic cycle that is energetically active). In older
and onset of the next inspiration. The parameters research, this parameter is referred to as I-frac-
most commonly used to describe breathing pat- tion (Woodworth and Schlosberg, 1955). Some
terns are depth and rate of breathing. The vol- of the older literature has also employed the
ume of air that is inspired or expired during one inspiration/expiration ratio (I/E ratio; Stoerring,
cycle of respiration is called tidal volume (Vt). 1906).
Respiration rate (RR) refers to the number of
respiratory cycles per minute. The duration of 3.2. Quantitative parameters
each respiratory cycle (Ttot), can be broken down
into inspiration time (Ti), and expiration time Gautier (1980) argues that mean inspiratory
(Tel. Sometimes, the pauses after inspiration (Pi> flow rate (Vt/Ti) may only be regarded as a valid
and expiration (Pe> are also included in the anal- measure of the drive component of the ventila-
yses. Traditional analysis of respiration only in- tory output, when the inspiratory volume-time
volves Vt and RR. However, the limitation of this profile (Vt/Ti) is linear. When the Vt/Ti rela-
type of analysis is that it does not provide insight tionship is curvilinear, the use of a shape factor
in the mechanisms which determine Vt and RR. should be considered. In this case Vt/Ti is no
For example, a change in Vt can be accomplished longer a pure index of inspiratory drive be-
by a change in the mean inspiratory flow rate cause it is time dependent. The same applies to
(Vt/Ti), by a change in Ti, or both. Similarly, RR mean expiratory flow rate (Vt/Te). A curvilinear
can be altered by changes in Ti or Te, or both. expiratory volume-time profile indicates that the
The breathing pattern appears to be regulated Vt/Te relationship does not reflect the actual
by at least two interacting mechanisms: a central rate of emptying of the lungs. For instance, the
inspiratory drive mechanism which is cyclically exhalation phase might include an expiratory
switched on and off by a timing mechanism in pause. Qualitative respiratory measures reflecting
the respiratory centre of the brainstem (Bradley, the morphology of the respiratory curves have
1981; Gautier, 1980; Milic-Emili and Grunstein, indeed been employed in a few investigations.
1976; Milic-Emili, Grassino and Whitelaw, 1981; Clausen (1951), for instance, examined the
von Euler, 1977). An enhanced inspiratory drive smoothness and the shape of the respiratory
primarily augments the mean inspiratory flow rate curve, and the angle of transition between inspi-
(Vt/Ti) with only secondary effects on timing ration and expiration. More recently, harmonic
(Milic-Emili et al., 1981). The contribution of analysis techniques for quantifying the shape of
drive and timing mechanisms upon ventilation the respiratory cycle have been described by
may be assessed by the following expression Bachy et al. (1986) and Benchetrit et al. (1989).
(Gautier, 1980; Milic-Emili and Grunstein, 1976; Quantitative analysis of breathing have also in-
Milic-Emili et al., 1981): cluded the abdominal-thoracic ratio (the ratio of
the amplitudes of thoracic and abdominal excur-
Ti
Vmin=RRXVt=GX- sions). Less attention has been devoted to sighing
Ttot and yawning, respiratory responses that are both
where minute ventilation volume (Vmin> refers to characterized by a long and deep inhalation. Pa-
the total ventilation during a 1-min period, mean rameters reflecting the time course of respiratory
inspiratory flow rate (Vt/Ti) is considered to be activity, such as regularity, duration of changes,
an index of the intensity of the central inspiratory systematic changes over time, have only been
 F.A. Boiten et al. /International Journal of Psychophysiology 17 (1994) 103-128 107

included in very few studies (e.g., Santibanez and (e.g. mouth piece, facemask, tubes with valves).
Bloch, 1986). Some researchers have assessed ir- Vt and RR can be greatly altered under the
regularity of rate and depth of breathing (e.g., influence of these devices (Askanazi et al., 1980;
Alechsieff, 1907; Hormbrey et al., 1986; Lehmann, Gilbert et al., 1972). A switch from nasal to oral
1914). breathing seems also an important determinant
of the change in volume and time components of
3.3. Gas exchange parameters the breathing pattern with a mouth piece and
nose clip (Perez and Tobin, 1985). In addition,
Under normal circumstances, ventilation is Perez and Tobin showed that alteration in
mainly regulated by arterial carbon dioxide (CO,) breathing pattern with a mouth piece and nose
tension. The arterial CO, tension represents the clip was associated with a significant reduction in
degree to which ventilation is in accordance with end tidal CO,, which persisted for at least 90
metabolic processes. Excessive ventilation (i.e., min. Furthermore, intrusive respiratory equip-
hyperventilation) causes more CO, to be elimi- ment may heighten the subjects awareness of his
nated from the body than is currently produced own respiration, which is known to alter the
by metabolic processes, and results in a drop in pattern of breathing (Western and Patrick, 1987).
arterial and alveolar pC0, below the normal Recently however, techniques have become avail-
range (i.e., 36-45 mm Hg; Bass and Gardner, able that enable non-invasive assessment of both
1985; Grossman and Wientjes, 1989). For safety time and volume components of respiration
reasons, arterial blood samples are usually not (Chadha et al., 1982; Cohn et al., 1982; Morel et
collected in psychophysiological investigations. al., 1983; Sackner et al., 1980; Wientjes, 1992).
Instead, the arterial CO, pressure is estimated by These techniques employ measurement of the
measuring the percentage or partial pressure of separate motions of rib cage and abdomen, either
CO, at the end of a normal expiration (end-tidal via inductive plethysmography (Watson, 1980) or
pCO,, Gardner et al., 1986). Measurement of via bellows pneumographs (Morel et al., 1983).
end-tidal pC0, is an important research tool to An essential element of these techniques is the
assess whether ventilation is in balance with calibration procedure that is used to estimate the
metabolical demands. Other gas exchange param- volume to motion coefficient (VMC) for the rib
eters include 0, consumption (VO,) and CO, cage and abdomen signals (Gribbin, 1983). Dif-
production @CO,). On the basis of these mea- ferent calibration techniques have been described
sures, energy expenditure may be estimated. Al- (Chadha et al., 1982; Gribbin, 1983; Morel et al.,
though this may be quite useful in certain respira- 1983) that employ a calibration session during
tory psychophysiological studies, very little (if any) which the subject either is breathing with a fixed
use has, to our knowledge, been made of these and known volume, or during which Vt is simulta-
measures in the context of emotion research. neously measured by means of a pneumotacho-
Therefore, they will not be discussed or referrred graph or a spirometer. The rib cage and abdomi-
to in this review. nal VMCs are estimated by means of multiple
regression. Unfortunately, attempts to calibrate
respiratory measurement devices are almost
4. Recording techniques nonexistent in most of the older psychophysiolog-
ical literature on respiration. Consequently, accu-
Recordings of tidal volume and respiratory rate measurement of respiratory volume is un-
rate can be obtained with a spirometer or common in this literature. In most cases, only
flowmeter (pneumotachograph). These recording inferential statements can be made about volume
devices directly measure the volume of inspired changes during emotional stimulation. In these
and expired air. This type of measurement is instances, we will refer to depth of breathing
obtrusive since it typically employs equipment instead of Vt.
that adds dead space and resistance to breathing The most commonly used gas exchange mea-
108 F.A. Boiten et al. / It~ternationul Joumal of Psychophysiology 17 (1994) 103-12X

sure in psychophysiological studies is the pC0, in excluded from the analyses. RR increased in both
end-tidal air. CO, measures are generally ob- fear and anger, but the increase was greatest
tained by a infrared gas analyzer. By means of a during anger. Schachter (1957) replicated the
heated sample tube attached to an open face study, adding pain as induced by the cold pressor
mask or to one of the nostrils, the CO, content of test. Fear and anger again yielded strong RR
the expired air is sampled and analyzed. increases, but the RR response during pain was
small. Fenz and Jones (1972) succeeded in induc-
ing a real life fear situation of an intensity not
5. Effects of negative emotions readily replicable in the laboratory. For both
novice and experienced parachutists RR mea-
5.1. Respiration rate and depth of breathing sures were taken during a jump sequence; from
the moment they arrived at the airport, to the
A number of studies have found increases in final preparations for the jump: the ride up to the
respiratory activity during the induction of nega- aircraft, reaching exit point, and after landing.
tive emotional states. Rehwoldt (1911) asked three Novice jumpers showed a continuous increase in
subjects to imagine emotions of their own choos- RR during the prejump sequence, from approxi-
ing. On the basis of self-reports, the emotional mately 17-25 cycles per minute. In experienced
states were classified as either positive or nega- jumpers, on the other hand, a similar increase in
tive affect, and as calm, excited, or tense affect. RR was noted early in the jump sequence (up till
Excited negative affects (anger, fear, indignation) 24 cpm), followed by a gradual decrease towards
resulted in increased RR and, in almost all in- the moment of the jump. Increases in RR during
stances, increased depth. Depth sometimes in- fear or anxiety have also been noted by Zucker-
creased to three times the baseline value (as man et al. (19681, and several others.
measured in millimetres pen deflection). Tense An alternative method to induce intense emo-
affects, whether negative (impatience, suspense) tions is hypnosis. Dudley (1964) induced depres-
or positive (hope) were associated with a higher sion, anxiety, anger, and pain (headache), as well
RR, but variable (sometimes decreased) depth. as exercise and relaxation by hypnotic suggestion.
One problem with this study is that the self-re- Suggestions of anger, anxiety and exercise were
ports which led to the affect classifications may followed by an increase in RR, Vmin and oxygen
have been influenced by the respiratory re- uptake, and a decrease in end-expiratory pC0,.
sponses. Lehmann (1914) did not replicate the Agreeing with Schachters (1957) finding, pain
effects of imagery of both unpleasant and pleas- only caused an increase in respiratory activity and
ant affect upon RR, but found that depth of hyperventilation when the headache induction
breathing increased during these conditions. Im- evoked overt anger or anxiety. Hypnosis was also
agery instructions of fear, anger, wonder, and used by Freeman et al. (1986) with patients who
pain led to strong depth and RR increases in a were considered to suffer from hyperventilation
study by Feleky (1914), that will be discussed in syndrome (HVS), and controls. Negative emo-
more detail below. tional arousal was induced by imagery meaningful
Somewhat more recent research corroborates to each individual subject (phobic objects, be-
that RR and/or depth increase occur during reavement experiences, imagined events, etc.).
negative affect states. Ax (1953) induced fear and During emotional arousal, controls and HVS pa-
anger in subjects by creating an atmosphere of tients showed a slight increase in RR, as com-
alarm and confusion due to a dangerous mal- pared to the baseline condition, and a larger
function of the recording equipment, and one of increase in RR as compared to hypnotically in-
irritation and anger by an incompetent and arro- duced calm and relaxation. Respiratory volume
gant polygraph-operator. Subjects whose emo- was not assessed in this study. However, because
tional response was not strong enough (as judged the HVS patients responded with a larger fall in
on the basis of interview and observation), were end-tidal pC0, than the controls (which reflects
 F.A. Boiten et al. /International Journal of Psychophysiology 17 (1994) 103-128 109

hyperventilation), it can be argued that the in- mura (1984) subjects had to judge music, RR
crease in respiratory volume must have been correlated negatively with ratings of that music as
greater among the patients than among the con- melancholical or dismal. Why RR is often
trols. not affected by stress films might be due to the
An increase in RR (and depth of breathing) fact that not only the intended emotion (anxiety)
combined with a decrease in pCOZ appears to be is elicited but other emotions as well, which in
a common response pattern to certain types of turn might have an opposite effect on RR. For
stress. A study by Suess et al. (1980) further instance, several films also induced depression
illustrates this. This study, in which RR and end- and sadness which have been associated with a
tidal pC0, were measured, assessed the effects decrease in RR (Rehwoldt, 1911). A more parsi-
of threat of electric shock via an intriguing exper- monious explanation is that film and imagery
imental protocol. The subjects performed a highly induced affect as well is often not intense enough
ambiguous perceptual judgement task, and were to provoke changes in RR. The cold pressor data
told that they would receive an electric shock indicate that increased RR is not a necessary
after making ten judgement errors. During the consequence of painful stimulation, although Vt
task, the experimenter falsely informed the sub- is in most cases significantly enhanced (see e.g.
jects about the errors that were made, until the Allen et al., 1986; Allen and Crowell, 1989). As
fatal tenth error was about to occur. In fact, noted before, an increase in respiratory activity
the task was terminated after nine errors, and no might only occur when the pain stimuli evokes
shocks were given. RR increased, and end-tidal overt anger and anxiety. On the other hand, a
CO, dropped to hyperventilatory levels both dur- decrease in RR in response to painful stimuli
ing an anticipatory pre-task period and during might point to a voluntarily induced slowing of
the task. As was the case in the study by Freeman RR in order to cope with the stress situation.
et al. (1986), Suess et al. (1980) assessed no Indeed, several studies have shown that under
volume measures; close inspection of the data, certain stress conditions slow breathing can re-
however, strongly suggests that volume must also duce subjective indices of anxiety (see e.g. Gross-
have increased substantially during the pre-task man, 1983), although cold pressor pain percep-
and task periods. Several other studies (Landis, tion may not be affected (Lane, 1980).
1926; Schnore, 19.59; Skaggs, 1930; Suter, 1912; In many of the studies mentioned, changes in
Svebak, 1982; Willer, 1975) also found an in- RR and depth of breathing appeared to go hand
crease in RR and depth, or an increase in Vmin in hand. However, this is certainly not always the
(Finesinger and Mazick, 1940) in response to case. An increase in RR in combination with
threat of electric shock or other painful stimuli. shallower breathing was observed by Alechsieff
Increases in RR and depth of breathing have, (1907) under tenseness, by Rehwoldt (1911)
however, by no means been found in all negative during tense affects, by Suter (1912) during
emotional conditions. For example, absence of anxious anticipation of electric shock, and by
RR effects during stressful films has been re- Lehmann (1914) in subjects reporting unpleasant
ported by Goldstein et al. (1965); Adamson et al. affect that they sought to resist. Taken together,
(1972); Koegler and Kline (1965) and by Cohen et these findings emphasize that negative affects are
al. (1975). Similarly, Rimm and Litvak (1969) not invariably tied to RR, depth, or volume in-
who instructed subjects to silently read sentences creases, and that RR and depth are not always
that had negative affective content, found no RR coupled. The variation in, and discrepancy be-
effects. RR decreases have furthermore been re- tween, RR and depth changes, may in part be
ported during imagery-induced wonder and pain due to a lack of adequate standardization of
(Feleky, 1914), after presentation of simple un- recording techniques and data treatment. Equally
pleasant sensory stimuli (Ruckmick, 1936) and in likely, however, they at least in part reflect pre-
response to the cold pressor test (Craig, 1968; cisely what Rehwoldt (1911), working within the
Allen and Crowell, 1989). In a study by Naka- framework of Wundts (1902) three-dimensional
model of affect, held them to reflect: differences times were longer and pauses shorter than during
between excited and calm, and between tense presentation of a neutral film; these findings were
and relaxed emotional states. Within the domain observed in the thoracic but not in the abdominal
of negative affect, RR and depth appear to re- channel. A reduction in post expiratory pause
spond to degree of excitement and degree of duration in response to negative film clips (Boitcn
tenseness of the individual, and to each of these and Gerritsen, in preparation) and mental arith-
differently. metic (Boiten, 1993a) appears to reflect moderate
Several distinct respiratory response patterns stress levels. The former investigation revealed
thus emerge from the data. For example, excited some strong negative correlation ratings between
affects, produced by the stress situation itself, negative film affect (fear, tension and involve-
may be primarily associated with rapid deep ment) and the duration of expiratory pauses. On
breathing, and tense, anticipatory affects with the other hand, very intense negative affect (hor-
rapid shallow breathing. We will try to further ror> induced in some subjects extremely long ex-
specify these patterns after discussing the changes piratory pauses, which lasted up to 15 s, before
found in positive emotional states and during commencing the next inspiration.
mental effort and attention.
5.3. Thoracic and abdominal breathing
5.2. I/E ratios and respiratory pauses
Discrepancies between thoracic and abdomi-
Older studies of respiration have included in- nal breathing have been discussed since the early
spiration-expiration balance (I/E ratio) among days of research into associations between respi-
their parameters. Feleky (1914) found longer in- ration and emotion (Bell, 1806). This pioneer
spiration than expiration durations (l/E ratios reported that excited negative affect resulted
> 1) in most of her subjects when they imagined in larger increases in thoracic than in abdominal
negative emotions, except hatred. Inspiration records; I/E ratio tended to increase thoraci-
dominance was moderate, though, except in tally, and decreased abdominally under those
wonder (which may be assumed to include conditions, as compared to neutral control con-
surprise) and fear. Similar observations were ditions. The same, to somewhat lesser extent,
made by Rehwoldt (1911) in negative affects, applied to calm negative affects (almost all of
both excited and calm ones, but only in thoracic these under imagined emotional conditions).
recordings. I/E ratios tended to be larger in what Landis (1926) induced severe emotional upset by
he classified as excited than in calm negative cumulative aversive manipulations. Subjects had
affects. Low I/E ratios during feelings of tense- to undergo periods without food (44-47 h) and
ness were found by Drozynski (1911). Increases of without sleep (36-38 h). Late in the afternoon of
I/E ratio from about 0.70 to about unity were the second day subjects were given severe electri-
found, however, among subjects that were lying cal stimulation, until they felt that they could not
during a fake court trial. This increase did not endure the pain any longer. This resulted, among
occur when the subjects made similar efforts that other things, in enhanced thoracic respiration,
could not be understood as hiding the truth (Be- whereas abdominal breathing decreased. Similar
nussi, 1914). The precise meaning of the findings effects were observed by Faulkner (1941) who
on I/E ratios is unclear, since the investigators reported that strong emotions can alter the range
did for instance not separate respiratory pauses of abdominal-diaphragmatic movements. Pre-
from inspiration and expiration proper. That sep- dominantly thoracic breathing was found by An-
arate analysis of respiratory pauses is a potential coli and Kamiya (1979) and Ancoli et al. (1980) in
useful endeavour has been shown by Cohen et al. subjects viewing an unpleasant film, whereas a
(197.5), who applied advanced component analysis pleasant film resulted in dominant abdominal
of the breathing cycle. They found that stress breathing. In a study by Svebak (19X2), increase
films had no effect on RR, but that expiration of thoracic movements occurred in subjects per-
 F.A. Boiten et ul. /International Journal of Psychophysiology 17 11994) 103-128 111

forming a task during threat of electrical stimula- component of the startle response. Blatz (1925)
tion, without a change in abdominal movements. placed unsuspecting blindfolded subjects in a
For a long time, dominance of thoracic breathing chair which suddenly tilted backwards. Immedi-
has been viewed as an important feature of the ately after falling, RR went down and inspiration
hyperventilation syndrome (Garssen, 1986; Lum, was lengthened. If the fall of the chair occurred
1976); however, this claim has never been sub- during an inspiration, that inspiration was pro-
stantiated by well-controlled research. All in all, longed, and when it occurred during expiration
the available information seems to suggest that the expiration movement stopped abruptly and
enhancement of thoracic relative to abdominal gave way to inspiration. More recently, Harver
respiration occurs during either unpleasant or and Kotses (1987) observed similar respiratory
tense affect, or during both. activity in subjects who received white noise at
intensities of either 80 or 110 dB during either
5.4. Irregularity of breathing inspiration or expiration. When stimulation oc-
curred during inspiration, the inspiratory move-
Pronounced irregularity was found in patients ment was accelerated thereby decreasing the du-
suffering from anxiety disorders, particularly when ration of stimulated period. When stimulation
angry (Stevenson and Ripley, 19.52; see below). occurred during expiration, a phasic inspiratory
Similarly, hyperventilation syndrome patients, movement evolved, and by doing so lengthened
who often are anxious, fearful, tense and con- the duration of the stimulated period. Analysis of
fused (see Grossman and Wientjes, 1989) have subsequent respiratory cycles revealed a transient
been reported to have greater irregularity of res- increase in rate and tidal volume. Also our own
piratory parameters than controls (Hormbrey et research (Boiten and Gerritsen, in preparation)
al., 1986). Irregularity of RR was also reported by corroborates phasic inspiratory shifts in breathing
Alechsieff (1907) and Lehmann (1914) under in response to startle. The startle stimulus con-
states of what they called excitement. Lastly, sisted of a particular scene of a stress film: an
Mador and Tobin (1991) found that noxious stim- unanticipated attack of a mad dog. The respira-
ulation produced by shining a bright light into the tory response again depended on whether the
subjects eyes increased the variability (coefficient startle stimulus occurred during an inspiration
of variation) of Vt, Ti and Te. or an expiration. During the inhalation phase the
startle induced an abrupt acceleration and
5.5. Phasic respiratory actirity deepening of inspiration. Occurred the startle
during expiration than that expiration was stopped
More or less pronounced phasic interruptions abruptly and was immediately followed by a short
of breathing generally occur upon presentation of and fast inspiration, thereby significantly prolong-
unexpected or novel stimuli. A catch of breath or ing the expiratory phase duration and hence the
arrested breathing is regarded as the respiratory total cycle duration. Also individual differences in
component of the orienting reflex (Barry, 1982). the appraisal of the stress film appears to influ-
Barry showed that respiratory pauses (phasic sus- ence the reportage and elicitation of a startle
pensions of breathing), were good indicators of response. Subjects who experienced significantly
stimulus novelty but not of stimulus intensity or more negative affect (emotional responders)
significance. Other researchers described the res- showed a strong tendency (p < 0.07) to report
piratory component of the OR either as a de- more often a startle response (11 out of 12 sub-
crease in RR and increase in amplitude or as an jects) than the non-emotional responders (14 out
initial pause followed by reduction in RR and of 25 subjects). They also showed a (non signifi-
amplitude (see Porges and Raskin, 1969). Unex- cant) prolongation of the expiratory phase (635
pected intense or aversive stimulation generally ms) of the respiratory cycle which was affected by
appears to produce an inspiratory switch in respi- the startle stimulus, relative to the expiratory
ration, which seems to reflect the respiratory startle phase of the non-emotional responders.
II? F.A. Roitm ct al. /Internutronul Journal uf Psychophysiology 17 (1994) IO.?-12h

Thus, it appears that the occurrence and magni- affects, and slow and shallow in the calm ones.
tude of the respiratory startle response is moder- The latter findings correspond well with the find-
ately affected by the level of negative affect a ings of later, better controlled studies. RR dc-
person may experience in response to film stim- creases were found, for instance, in relaxation
uli. These results seem in line with recent re- (Skaggs, 1930; Dudley, 1964), and daydreaming
search suggesting that the vigor of the startle (Corwin and Barry, 1940). In Nakamuras (1984)
response varies systematically with the persons study of respiratory reactions in subjects listening
emotional state (Lang, Bradley and Cuthbert, to music, RR was lower as ratings of calmness
1990). Lastly, sudden cooling of large and sensi- were higher. A number of other studies have
tive body areas (e.g. chest and abdomen) has also observed no effects of pleasant stimulation upon
been shown to induce a abrupt transient inspira- RR. Finesinger and Mazick (1940) reported only
tory shift (gasp> in breathing (Keating and Nadel, very small changes during pleasant stimulation.
1965; Mekjavic et al., 1987), followed by an imme- In a review article on physiological measures of
diately increase in RR and depth of breathing, sexual arousal Zuckerman (1971) discussed sev-
and a fall in end-tidal pCOz (Cooper et al., 1976). eral studies dealing with respiration and sexual
An initial gasp accompanying startle has further- arousal. He concluded that gentle sexual arousal,
more been reported by Hogan (1970). in most studies induced by visual stimuli, had
Thus, unexpected aversive stimulation pro- only marginal effects on RR: Apparently, respi-
duces two phasic respiratory responses. A short- ratory measures will not be useful in assessing
latency respiratory startle response, followed by a sexual arousal to visual (sexual) stimuli (Zucker-
delayed phasic increase in depth and rate of man, 1971, p. 309).
breathing. In addition, the inspiratory gasp has a On the other hand, some studies have de-
short onset latency which is in line with the fast scribed an increase in RR during positive affect.
rise time of other indices of the startle response, In some instances, calm pleasant affects were
such as heart rate acceleration and cephalic va- in Rehwoldt (1911) study associated with in-
sodilatation (Turpin, 1986). Turpin also argued creased RR and decreased depth (fast and shal-
that the delayed, or long latency increases in low breathing), instead of the slow and shallow
physiological activity reflect fight and flight re- breathing that was generally found. Ruckmick
sponding. (1936) found the same fast and shallow pattern
among subjects watching simple pleasant stimuli,
such as slides of faces with emotional expressions.
6. Effects of positive emotions Lehmann (1914) generally observed both RR and
depth increases in (imagery provoked) pleasant
6.1. Respiration rate and depth of breathing affects. Nakamura (1984) found that music rat-
ings of cheerful, gay and powerful corre-
Psychophysiologists have paid almost no atten- lated positively with RR. Strong increases in RR
tion to the effects of pleasant stimulation or have been reported to occur during coitus by
positive emotions upon respiration. The studies Masters and Johnson (1966) and by Bartlett
that do exist, however, are generally consistent (1956). Masters and Johnson (1966) reported that
with our earlier suggestion that changes in RR hyperventilation developing during the late
and depth are more strongly associated with plateau and the orgasmic phases of the sexual
Wundts (1902) excitement/ calm and tense/ response cycle is a normal occurrence. RRs of 40
relaxed dimensions, than with pleasantness as c/mm have been recorded repeatedly during or-
such. In many studies of positive affect, RR and gasm (both sexes). Bartlett (1956) reported
depth decreases were found. In Rehwoldts (1911) marked peaks at orgasm with RRs of 20-70/ min.
study, decreases in RR and depth occurred dur- Most findings in this modest collection fit in
ing positive affects induced by imagery. Respira- the patterns that we mentioned earlier: an in-
tion was slow and deep in the excited positive crease of RR and volume during excitement, and
 F.A. B&en et al. /Internationul Journal of Psychophysiology 17 (1994) 103-12X 11.3

a decrease in volume, combined with an increase correlations with thoracic amplitude, were found
in RR, during pleasant attentive states. The ma- in men.
jor deviation from those patterns is formed by the
excited affects in Rehwoldts old study, during 6.3. I/E ratios and respiratory pauses
which a pattern of slow and deep breathing was
found. This suggests an activated (rather than Rather variable I/E ratios have been reported
aroused) respiratory response type, that is, a in positive affects. Calm positive affect, in Re-
response type corresponding to tonic readiness to hwoldts (1911) study, tended to yield longer expi-
act rather than to sudden, reactive, response mo- rations than inspirations (I/E ratios < 1). Ex-
bilization (see Pribram, 1981). cited positive affects showed ratios > 1 for tho-
racic, and < 1 for abdominal breathing. Feleky
6.2. Thoracic and abdominal breathing (1914) reported inconsistent results for imagining
pleasant situations. I/E ratios in laughter, of
Under positive emotional conditions, tho- course, were very low (average of 0.52 in Felekys
racic/abdominal ratio has been observed to shift 6 subjects).
in the direction of stronger abdominal than tho- The duration of expiratory pauses seems a
racic breathing amplitude. Faulkner (1914) re- stable characteristic of rest or relaxation, dimin-
ported that when subjects were asked to imagine ishing whenever breathing is stimulated (Newson
pleasant situations, the amplitude of the abdomi- Davis and Stagg, 1975). This is in agreement with
nal movement increased. Ancoli and Kamiya strong positive correlation ratings between sub-
(1979) and Ancoli et al. (1980) found dominant jective ratings of relaxation and the duration of
abdominal breathing in subjects viewing pleasant expiratory pauses (Boiten and Gerritsen, in
films. In the calm positive affects in Rehwoldts preparation).
(1911) study, abdominal breathing was also domi-
nant. Abdominal dominance is not found under
all more or less pleasant conditions. Rehwoldts 7. Mental effort and stress
(1911) excited positive affects yielded, on the
whole, equal abdominal and thoracic amplitude. Several researchers have made attempts to as-
Timmons et al. (1972) studied abdominal-thoracic sess the effects of stressful mental load tasks
movements in subjects who were instructed to upon respiration. Extensively studied mental load
relax and try to go to sleep. Abdominal-dominant tasks are mental arithmetic and reaction time
breathing was associated with relaxed wakeful- tasks, performed under stressor conditions such
ness, equal to drowsiness, and thoracic-dominant as task difficulty (Svebak, 1982; Carroll, Turner
breathing with sleep onset. and Hellawell, 1986), thread of electric shock
Care in the analysis of thoracic/abdominal (Allen et al., 1986; Sherwood et al., 1986) and
balance is needed in light of sex differences that making money rewards contingent on task perfor-
have been reported. According to Clausen (1951), mance (Turner and Carroll, 1985; Allen and
women are relatively stronger thoracic breathers, Crowell, 1989). Some of the earlier researchers
whereas men appear to be predominantly abdom- (Alechsieff, 1907; Suter, 1912; Skaggs, 1930)
inal breathers. These dominance differences have showed that mental arithmetic produced shallow
consequences for the leeway remaining for emo- breathing, with RR either not changing or in-
tional response changes. Svebak (1975) found that creasing marginally. Decrease in depth and vari-
in women duration and frequency of laughter in ability of the breathing pattern was also observed
response to funny films are strongly correlated during concentrated sensory attention, presum-
with abdominal (and not with thoracic) ampli- ably serving to suppress distracting and irregular
tude, both during the laughter and the silent breathing movements, which might otherwise in-
periods of the entertainment. Zero correlations terfere with task performance (see also Suter,
with abdominal amplitude, and non significant 1912; Woodworth, 1938; Kagan and Rosman,
113 F.A. Boiten et al. /International Journal of Psychophysiology 17 (1994) 103-128

1964). Recent research generally supports these components, and, consequently, in RR. Com-
early findings. For instance, Carol1 et al. (19X6) pared to pre-task resting levels, Vt decreased
engaged subjects in mental arithmetic and during the least stressful condition, but the de-
Ravens matrices. Each task was presented with crease in Vt became less pronounced as the task
three levels of difficulty: easy, hard and impossi- demands increased, and reached, during the most
ble. In addition, subjects were told that appropri- stressful condition, a level that was nearly equal
ate responses would be financially rewarded, to the initial resting value. Both the timing and
whereas erroneous responses would lead to a the drive mechanisms of respiratory control
reduction of the monetary rewards. Analysis re- (Gautier, 1980; Milic-Emili and Grunstein, 1976;
vealed that an increase in task difficulty covaried Milic-Emili et al., 1981) turned out to be affected
with a significant increase in RR, a non-signifi- by the stress level: Vt/Ti increased from rest to
cant decrease in Vt and a significant increase in task, and increased further as the level of stress
average ratings on an arousal scale (calm/ at ease increased. Ti/Ttot only increased in the most
vs excited/ aroused), indicating that for both tasks stressful task condition. Thus, in the Wientjes
the hard and impossible condition were perceived (1993) study, the respiratory pattern during
as more arousing than the easy condition. That stressful mental task performance was character-
enhanced arousal levels increase RR has also ized by relatively fast, shallow breathing, with a
been reported by Schnore (1959), who induced high mean inspiratory flow rate. As stress levels
moderate arousal levels in his subjects by visual increased, the breathing pattern became less shal-
pursuit tracking, and high arousal by threatening low, and mean inspiratory flow rate was even
them with strong electric shock if they failed to further augmented, but there was no change in
improve the best score that they had previously the rate of breathing.
obtained. An increase in RR was found during The respiratory changes during mental effort
the high arousal, as compared to the low arousal thus seem to resemble the pattern that was ob-
condition. Several other studies have produced served during tense affects, during anxious an-
evidence that a range of stressful laboratory tasks ticipation of an aversive stimulus, and among
elicit an increase in RR, and a slight decrease in patients with anxiety states (see below): rapid
Vt (Allen, Sherwood and Obrist, 1986; Carroll et shallow breathing. This similarity can be ex-
al. 1987; Langer et al, 1985; Sims et al., 1988; plained in several ways. Both the performance of
Svebak, 1982; Turner and Carroll, 1985; Turner, mental tasks and states of tension and anxiety
Carroll and Courtney, 1983; Wientjes, 1993; may well involve sustained attention; also, both
Boiten, 1993b). Since increases in RR are on the involve voluntary or inhibitory response control.
whole more pronounced then decreases in Vt, Another similarity is that most mental tasks, like
Vmin often is augmented. states of tension and anxiety, include a stress
Wientjes (1993) measured additional respira- component (such as task difficulty, time limits,
tory parameters during stressful mental load tasks, and positive or negative rewards).
via inductive plethysmography, and performed a
computer based component analysis on the cali-
brated respiratory signal. There were three reac- 8. Specific respiratory responses: sighing and
tion time task conditions in the study, that were yawning
equally difficult, but that differed in the degree to
which they were effort-demanding and stressful. 8.1. Sighing
In all three conditions, duration of the time com-
ponents of the breathing cycle (Ti, Te, pause A sigh is a single breath characterized by a
times, and Ttot) decreased (and hence RR in- relatively long duration, deep in- and expiration,
creased), as compared to the pre-task resting and a particular sound. In a sigh, a large volume
condition. Between conditions, however, there of air is inhaled. However, the air passages are
were no differences in the duration of the time not widened during inspiration and remain tight-
 F.A. Boiten et al. /International Journal of k7ychophysiolog.v 17 (1994) 103-128 115

ened in the expiration phase. The discordance yawning is a muscular and vascular reflex, the
between the volume of air that is displaced and physiological function of which might be to im-
the calibre of the airways suggests that the sigh prove circulation and to stretch the musculature.
does not serve a need for increased oxygenation As to its psychological aspects, it has been pro-
but is the expression of an emotional state (Bend- posed that yawning is an expression of an un-
ixen et al., 1962; Harrer, 1969). Sighing has been pleasant state, namely boredom (Provine and
observed in nervous, fatigued, bored and excited Hamernink, 1986). However, yawning is certainly
persons (Cabot and Adams, 1942). Finesinger not only tied to unpleasant affect: its occurrence
(1939) reported that the induction of unpleasant may also be associated with pleasant, comfortable
ideas is associated with an increase in sighing. feelings. Lehman (1979) argues that yawning is a
Sighing has also been related to anxiety (Steven- signal that a person is making an effort to main-
son and Ripley, 1952; Tucker, 1963). Tobin et al. tain contact with the outside world. Others have
(1983) reported that patients with chronic anxiety argued, from an evolutionary point of view, that
states exhibited sighs at a rate of 4-2.5 over a 15 yawning is part of a silent sign language, and
min monitoring period. In young normal subjects may serve to express the friendly or non-hostile
the frequency of sighing was 0-l sigh per 15 min. intentions of an animal. Experimental evidence is
Clausen (1951) reviewed some studies on respira- as far as we know not available.
tion in neurotics, from which he concluded that
sighing is the symptom most generally observed in
these subjects. These studies imply that sighing 9. Differentiating between emotions
mostly occurs during unpleasant states, except
perhaps for sighs of relief. Garssen (1986) hy- As far as we know, only two studies have
pothesized that the function of sighing might be systematically explored differences in respiration
the relief of feelings of tightness around the chest, between specific emotions in normal subjects.
and that this could explain why anxiety states are Feleky (1914) induced what she considered to be
often associated with sighs. However, Garssens the six primary emotions (pleasure, pain, anger,
attempts to confirm this hypothesis failed. The wonder, fear and disgust) by asking subjects to
psychological function of sighing remains there- relive a past emotional experience. Measures of
fore unclear. respiration were I/E ratio, depth of breathing,
and the amount of respiratory work: i.e. depth
8.2. Yawning divided by breath cycle duration. RR can be
derived from the variables work and depth,
There is remarkably little physiological or psy- Table 1 reproduces her average data. We per-
chological literature available on the subject of formed four one-way variance analyses (one for
yawning. Yawning is characterized by a pro-
longed inspiration with wide open mouth, fol- Table 1
lowed by shorter expiration and is often associ- Emotion patterns according to Feleky (derived from Table 2,
ated with shivering and stretching movements of Feleky, 1Y14)
the muscles of the trunk and extremities. One of Emotion l/E Work Depth Rate
the most remarkable aspects of yawning is its disgust 1.1 0 5.9 0 15.3 0 23.4 +
contagiousness; something which has also been pleasure 1.1 0 Y.5 + 27.3 + 20.4 +
observed among animals. Like in sighing, the anger 1.5 + 13.3 + + 31.8 + + 25.2 + +
pain I.5 + 8.0 + 20.7 + 16.2 0
deep inspiration would suggest that yawning may
wonder 23 + + 9.5 + 38.6 + + Y.6 0
serve to increase the bodys oxygen uptake. How- fear 2.6 + + 14.6 + + 33.2 t + 76.4 + +
ever, a yawn is always followed by a period of normal 0.8 0 4.7 0 13.9 0 20.4 +
apnoea, so nothing is gained as far as oxygen
See text for explanation of the respiratory parameters. Zeros
metabolism is concerned (Lehman, 1979). Ac- and plus signs indicate three crude classes of response magni-
cording to Mayer (19211, cited by Lehman (1979), tudes within each response variable.
each parameter) on the individual data from the pattern for joy was much the same as that of
six emotions and normal. and four others from sadness, except that the high-frequency oscilla-
the six emotions only. All but one were significant tions primarily occurred during the cxpiratory
at p < 0.0001 (the only exception was at p = phase. High-frequency movements during laugh-
0.0016), indicating that the emotions differed on ter appear to correspond to the ha-ha re-
all four respiration parameters. Furthermore, if sponses (Svebak, 197%. Tenderness was associ-
the averages are divided into three crude classes ated with a decrease in RR below the baseline
of response magnitude, all seven patterns turn value, and an increase in depth, yielding a pat-
out to be distinctly different, as can be seen in tern of slow and deep breathing. Again, the re-
Table 1. Of course, these data should be inter- sults are only indicative, and suffer from some of
prctcd with great caution. Subjects attached to a the same shortcomings as those of Feleky. No
pneumograph asked to relive emotions can hardly quantitative analyses were performed. Yet, the
bc said to have been naive. They may have been study merits attention, if only because it is one of
producing stereotyped response patterns, as these the few studies that have assessed the time course
also appear from subjective report studies (Rim& of respiratory responses. In a follow-up study
et al., 1990). Still, we have reported the data in (Bloch et al., 1991) young actors were instructed
some detail because they show the differentiation to voluntarily reproduce these six emotion-specific
potential in respiratory patterning. Moreover, respiratory patterns, together with the corre-
subsequent research has neither replicated nor sponding facial and postural configurations. A
rejected these findings. qualitative analysis of the recordings showed that
Marc recently, Santibafiez and Bloch (1986) as the emotional reproduction went along, both
observed specific respiratory (in addition to pos- breathing and facial expressions evolved from an
tural and facial) response patterns during emo- initial robot-like phase to a more natural stage
tional recall and revival in awake and hypnotized in which spontaneous vocalizations and gestures
subjects, that were associated with particular appeared. According to Bloch et al. this sug-
emotions. The emotions were: joy-laughter, sad- gested that the production of specific emotional
ness-crying, anger, fear, erotic arousal, and ten- configurations initiated the corresponding emo-
derness. Time and volume components, and mor- tional experience, although subjects were not
phology of the breathing cycle were compared questioned to confirm this. Respiratory analyses
between baseline and emotion states. With the were conducted during the emotional reproduc-
exception of tenderness, all emotions evoked a tion phase which followed the initial robot like
fairly irregular respiratory pattern. Characteristic phase. The results showed that the derived
for fear and anxiety were shallow breathing or breathing patterns were significant different for
cessation of breathing, and an irregular RR. the six emotional configurations. Relative to a
Anger was associated with an increase in RR neutral condition, the tenderness configuration
without changes in depth, but only at the onset was characterized by a slight increase in expira-
and at the end of the emotion induction. In the tory pause duration, erotic arousal showed a sig-
intermediate period, RR did not further increase, nificant decreae in expiratory pause duration,
but depth increased to three times base value, in whereas anger breathing was characterized by
a waxing and waning pattern. Erotic arousal (in- strong increases in rate and depth of breathing.
duced by making the movements of sexual re- Both the joy-laughter and sadness-crying configu-
sponse) was characterized by an initial decrease rations showed significant increases in depth of
in RR, followed by an increase in RR and depth. breathing while rate decreased, also high-
Sadness was also associated with a very character- frequency oscillations were superimposed on the
istic response pattern. RR remained stable but breathing curve. Lastly, the fear configuration
depth distinctly increased, as compared to the was characterized by substantial increases in
baseline period. Also, high-frequency oscillations depth of breathing, inspiratory shifts in end-ex-
were superimposed on the inspiratory curve. The piratory lung volume, and a total absence of
 F.A. Boiten et al. /International Journal of Psychophysiology 17 (1994) 103-128 117

expiratory pauses. Bloch et als (1991) claim, that 1935; Coppen and Mezey, 1960; Finesinger, 1943;
the production of emotion-specific respiratory-fa- Stevenson and Ripley, 1952). Though our main
cial-postural configurations can initiate emotional interest is in breathing patterns recorded among
experience, is intriguing and obviously has impor- normal subjects, we will discuss the most impor-
tant implications for the hypothesis that emo- tant clinical studies concerning anxiety neuroses
tional states are associated with specific physio- and depression. It is in these studies that investi-
logical responses. Unfortunately, the experimen- gators claim that there is e.g. hardly an emotion
tal support is not as strong as suggested. For which has not its obvious respiratory manifesta-
instance, there is as yet not much evidence that tion (Christy, 1935) or respiratory patterns vary
the voluntarily reproduced breathing movements closely with the emotional state (Stevenson and
correspond to emotion-specific breathing pat- Ripley, 1952).
terns. Also, data on emotional self-report is lack- There is considerable speculation about the
ing and the number of derived breathing curves is role of respiratory influences in the aetiology of
often too small to allow for sound analyses. In psychosomatic disorders such as the hyperventila-
addition, the reported respiratory differences tion syndrome (HVS), panic disorder (PD) and
might be due to the differences in effort in pro- agoraphobia. A common factor in the clinical
ducing the various configurations; for instance, a features of these disorders is the combined expc-
fearful expression is more difficult to produce rience of anxiety and somatic symptoms. On the
than a happy expression and consequently has a other hand, their clinical manifestations are
greater impact on breathing (Boiten, submitted). somewhat dissimilar: patients suffering from PD
In Axs (19.53) study, respiration rate increases and agoraphobia typically present with attacks
were greater for fear than for anger. Depth in- that are characterized by a sudden onset of symp-
creased as compared to the control condition, but toms and by abrupt, uncontrollable panic, whereas
there was no difference between anger and fear. abruptness of attacks is not considered to be a
In Schachters (1957) replication, fear and anger necessary feature for the diagnosis of HVS. Al-
produced the strongest increase in RR, while though there is some variation in the aetiological
pain induction led to only a slight RR increase. models that have been proposed by different au-
The patterns during imagined anger and fear thors, several investigators have hypothesized that
were similar to those found by Ax (1953) during the symptom attacks that typify these disorders
real emotion induction. are triggered by a distinct sequence of events. In
Emotions with a specific respiratory feature this regard, it is commonly assumed that anxiety-
appears to be startle and surprise, which are induced hyperventilation produces somatic symp-
characterized either by an inspiratory shift in toms which are, in turn, cognitively misinter-
breathing (startle) or a brief suspension of breath- preted in a catastrophic manner and thereby cause
ing (surprise). The phenomenon of respiratory an increase in anxiety. This leads to a vicious
suspension fits nicely into the interpretation of spiral, in which further increases in anxiety result
surprise as in part being an inhibitory response in further respiratory increases and in more in-
patterns (orienting response; Sokolov, 1963; tense symptoms (e.g. Clark et al., 1988; Ley,
movement arrest and jaw muscle relaxation; Du- 1985). Although there is evidence that excessive
mas, 1948). ventilation and hypocapnia may indeed, to a mi-
nor degree, be involved in the formation of so-
matic symptoms (Wientjes, 19931, empirical sup-
10. Studies among clinical populations port for hyperventilation models of HVS, PD and
agoraphobia has by no means been unequivocal
In most studies that were performed in clinical (e.g. Buikhuysen and Garssen, 1990; Grossman
settings, the respiratory pattern was used to dif- and Wientjes, 1989; Hornsveld et al., 1990). As an
ferentiate patients from normal subjects alternative, it has been proposed that the symp-
(Alexander and Saul, 1940; Burns, 1971; Christy, tom attacks may be due to enhanced tendencies
of anxious and distressed individuals to focus Clinical depression, on the other hand, ap-
their attention upon bodily sensations, and to pears to be associated with increased RR, with
appraise these in a negative manner (Watson and decreased levels of pCO,, and with diminished
Pennebaker, 1989). Notwithstanding these theo- CO, sensitivity (Damas-Mora et al.. 1976;
retical considerations, thcrc is ample evidence Damas-Mora et al., 1982; Shershaw et al., 1073;
that clinical anxiety disorders may be associated Shershaw et al., 1976). At present, it is unclear
with aberrant respiration. whether the respiratory characteristics of clinical
Respiratory complaints that are often associ- depression are associated with the depressed af-
ated with anxiety disorders are breathlessness fects as such, or rather with the anxiety compo-
and dyspnoea (difficult or laborious breathing). nent that is often observed among clinically de-
Patients typically complain of air hunger (a pressed patients (Watson et al., 1988; American
subjectively experienced inability to get enough Psychiatric Association, 1980).
air in the lungs), and a persistent feeling of pres- Stevenson and Ripley (1952) evaluated respira-
sure or heaviness on the sternum. Breathlessness tory patterns of patients with asthma and anxiety
and dyspnoea appear to be tied to hyperventila- states during relaxation and during episodes of
tion and sighing (Burns, 1071; Grossman and emotional disturbance, such as anger, anxiety,
Wientjes, 1989). Christy (lY35) found that the guilt and depression. Since these authors investi-
respiratory pattern of patients with anxiety disor- gated a broader range of emotions than is com-
ders who complained of breathlessness was char- mon in the literature, we will review their study
actcrized by a tendency towards rapid and shal- in some detail. Emotions were induced by having
low breathing, and by irregularities of depth of the patients think and talk about various cmo-
respiration. Finesinger (1943) and Tobin et al. tional experiences relevant to their life and ill-
(1983) also reported sighing and irregular breath- ness. The patients emotional statements were
ing during chronic anxiety (episodes of rapid shal- related to the observations of the physician with
low breathing alternating with episodes of slow, regard to voice intonation, facial expression,
deep breathing and transient inspiratory shifts in movements of the hands or body, and other in-
end-expiratory lung volume). Shallow breathing. dices of emotional change. Respiratory patterns
and chronic inspiratory shifts in breathing caused were examined with respect to RR, depth, I/E
by the incapacity to complete a full expiration, ratio and irregularities. Increases in RR and depth
were also observed by Reich (1949) in tensed and of breathing wera found chiefly during acute anx-
anxious patients. Hormbrey et al. (1986) found iety, and sometimes during anger and resent-
HVS patients (who typically are tense and anx- ment. Decreases in respiratory function where
ious and suffer from various psychosomatic com- found when the patients felt tense and on guard
plaints) to breathe more irregular than controls. with feelings of anxiety or anger and when feeling
Similar findings arc reported in two studies cited sad or rejected. Irregularity of breathing was not
by Bass and Gardner (1985). Jones and Scaris- primarily associated with acute anxiety. but rather
brick (1941) and Friedman (1045) found that pa- with anger.
tients with anxiety disorders had a higher RR, a Stevenson and Ripley interpreted their find-
smaller Vt and shorter breath-holding times than ings in terms of adaptation of the respiratory
controls. Thus, rapid and shallow breathing, sigh- system to the type of stress. They argued that the
ing and irregular breathing patterns seem impor- nature of the subjects stress response dctcr-
tant features of respiration in patients suffering mined whether a decrease or increase in rcspira-
from anxiety disorders (see also Baker, 1934; tory function would occur. Thus, they took an
Malmo and Shagass, 1949; Finesinger, 1943). De- increase in RR and depth to reflect physical
pressed affect, sadness and misery appears to be adaptation to stress (preparation for fight or
associated with the opposite respiratory pattern. flight), and suggested that a decrease in respira-
During these emotions, a decrease in respiratory tory function was associated with a tense expecta-
activity has been observed (Averill, 1969). tion in preparation for action. Stevenson and
Ripleys experimental design is an appealing ap- very small numbers of subjects, have employed
proach to the problem of examining emotion deficient experimental control procedures, inadc-
specific respiratory patterns. An interview context quate measurement techniques, or (at best) prim-
allows for a more precise evaluation of the sub- itive statistical treatment of the data. Although
jects prevailing emotional state, which in turn the more recent literature generally has a higher
can be synchronized with respiratory function. In level of methodological sophistication, there arc
this way one can control for feelings of hostility only few studies available that may be regarded
and tension evoked by experimenter or experi- as methodologically sound. In this light, it is
mental setting. An experimental interview context difficult to draw specific and detailed conclusions
might also allow for control over additional emo- concerning the influence of emotions upon respi-
tional behaviour. For instance, anxious patients ration. This does not imply that the available
are often also depressed, and most depressed literature is insignificant or irrelevant. To the
patients exhibit an elevated degree of anxiety contrary, this literature has produced many sug-
(Zuckerman et al., 1968). Without interview data, gestive results, that show an appreciable amount
it might prove difficult to separate anxiety and of consistency across studies. Consequently, sev-
depression as distinct affects (see Watson, Clark eral strands of evidence may be discerned, that
and Carey, 1988). This mixing of emotional run across different studies. On this basis, we will
behaviour might explain why breathlessness has try to draw some tentative general conclusions
also been observed in depression (Burns, 1971). It that may be of value for future research. More
should however be noted that the relationship specifically, with respect to the most widely em-
between emotions and respiration might have ployed respiratory variables, RR and depth, sev-
been confounded by the metabolic requirements eral distinctive patterns of response can be distin-
of the patients conversation. It is known, for guished. These appear to correspond in a more
instance, that the act of speaking itself is suffi- or less orderly fashion to general dimensions of
cient to cause elevations in heart rate and blood emotional and affective reactions that cut across
pressure (Lynch et al., 1980). Also, analysis of the the usual emotion boundaries, and even across
respiratory signal during speech in terms of vol- the distinction between pleasant and unpleasant
ume and time components is notoriously difficult emotional states.
because of the irregular expiratory movement of
air through the glottis that is typical of voice 11.1. Fast and deep breathing
intonation (Wientjes, 1993).
Most of the preceding studies seem to agree An increase in RR combined with an increase
that the respiratory pattern associated with anxi- in depth of breathing seems to be associated with
ety consists of irregular rapid, shallow breathing states of excitement. This is the generalization
with many sighs. The combined decrease in depth, offered by early researchers like Lehmann (1914)
and increase in rate of breathing that appears to and Rehwoldt (1911), and by Woodworth (1938;
be typically observed among patients with anxiety Woodworth and Schlosberg, 1955). More recent
disorders again fits in well with the pattern that is research supports this hypothesis. By excite-
found during tense affects and during the perfor- ment we mean the state that generates undi-
mance of mentally demanding tasks. rected, aimless behaviour, under conditions in
which directed action is blocked or restrained.
Feelings of excitement can be understood as the
11. Discussion felt urge towards such undirected action (Frijda,
1986; Frijda et al., 1989). Excitement, in this
As is apparent from the previous sections, the sense, often is an aspect of states of fear, anger,
literature on emotion-respiration relationships is and joy; but is not invariably so, because it can
fraught with methodological shortcomings. Many exist without any of these, as diffuse excitement.
studies (especially the older ones) have included We prefer the term excitement to that of
120 F.A. Boiten et al. /International Journnl qf Psychophysiology 17 (1994) 103-128

arousal, because the concept of arousal has during sustained attention (Kagan and Rosman,
many meanings and implications of generality 1964), and during tense or anxious anticipation or
that do not necessarily apply to excited emotions expectation (Skaggs, 1930; Stevenson and Ripley,
(and that, in fact, render the concept useless in 1952; Suter, 1912). Behaviourally, fast and shal-
discussing emotional states; see Neiss, 1988, 1990; low breathing appears therefore to be associated
Venables, 1984). Moreover, excitement, as we with (tense) anticipation, involving alertness and
defined it, is linked to an enhanced readiness for often attention, or with consummatory behaviour
action. Thereby, a meaningful connection can be characterized by some degree of control over
made with the primary function of enhanced res- effortfull action tendencies. In this sense, the
piratory activity. Such activity is usually inter- pattern seems typical for attempts to tune in to
preted as a preparation for flight or fight (Darwin, behavioural demands that may necessitate a re-
1872; Cannon, 1929; Stevenson and Ripley, 1952; strained, precise, and goal-directed (as opposed
Suess et al., 1980). Dudley et al. (1964, 1969) to massive) type of physical action. Emotionally,
indeed argue that changes in ventilation are pri- this pattern may apparently be tied to a range of
marily related to action or non-action orienta- affects varying from concentration, resolution and
tions. On the one hand, action-oriented respira- determination, through tension, to anxiety and
tory hyperfunction may be linked to emotions like fear, and maybe ultimately, panic. As we have
anger and anxiety. On the other, it may reflect a suggested before, this range of emotional re-
readiness for active appetitive behaviour in ex- sponding seems to correspond to a continuum of
cited positive emotions, or for physical exercise. respiratory patterns that extends from rapid shal-
This latter suggestion is supported by the obser- low breathing to rapid deep breathing. In this
vation that RR and depth increase have not only respect, the study by Wientjes (1993) is illustra-
been shown to occur during actual physical exer- tive. He has shown that an increase in mental
cise (Kay et al., 1975a, 1975b), but also during task stress produced a shift from shallow fast
hypnotic suggestions of such exercise (Morgan, breathing to deep fast breathing (whereby RR
1985). remained high, and changed very little). Hence, it
seems that variations in depth during rapid
breathing correspond with different points on a
11.2. Fast and shallow breathing inhibition-excitation continuum (which may be
more or less equivalent to a continuum extending
Rapid shallow breathing is certainly one of the from behavioural control to loss of control). In
most extensively documented breathing patterns. other words, the relatively shallow breathing pat-
As is suggested by the Wientjes (1993) study, tern that is usually found during states of concen-
mean inspiratory flow rate tends to be high in this trated attention, can apparently be modified by
pattern, which signifies that inspiratory drive is influences related to loss of control or increase in
high. When contrasted to the pattern of fast, emotional excitement.
deep breathing, this pattern appears to blend
readiness for action with some degree of in-
hibitory control. As we have mentioned earlier, 11.3. Slow and deep breathing
fast and shallow breathing is found during effort-
full and stressful mental task performance (Allen A pattern of relatively slow and deep respira-
et al., 1986; Carroll, Turner and Hellawell, 1986; tion is most often encountered in relaxed resting
Carroll et al., 1987; Langer et al., 1985; Sims et states (Wientjes, 1993; Boiten, 1993b). The pat-
al., 1988; Svebak, 1982; Turner and Carroll, 1985; tern tends to become slower and deeper during
Turner et al., 1983; Wientjes, 1993; Boiten, slow-wave sleep (Snyder and Scott, 1972; Douglas
1993b), in tense affects (Alechsieff, 1907; et al., 1982), and during relaxation (Grossman,
Lehmann, 1914; Rehwoldt, 1911), when a subject 1983). Other links between this pattern of breath-
is resisting unpleasant affect (Lehmann, 19141, ing and behavioural, or emotional responses have
 F.A. Boiten et al. /Intrmatior~al Journal qf Psychophysiology 17 (1994) 103-128 121

been reported in some of the older studies. It was to be characteristic for behavioural conditions
observed during excited positive affect (Re- that may either involve withdrawal from the envi-
hwoldt, 1914) and when the subject abandoned ronment, relaxation, or active coping (Allen et al.,
him- or herself to negative affect (Lehmann, 1986; Langer et al., 1986b; Wientjes, 1993). Hy-
1914). These data are insufficient to draw solid perventilation, on the other hand (which is char-
conclusions concerning the behavioural and emo- acterized by a drop in pC0, below the normal
tional correlates of this type of breathing. They range), appears to be the typical respiratory re-
do, however, suggest that the pattern might occur sponse during psychological challenges where very
both during relaxation as well as during unre- few, if any, active coping possibilities exist, such
strained emotional readiness to act. as states of threat, aversive stimulation or pain
(Allen et al., 1986; Dudley et al., 1964; Freeman
et al., 1986; Suess et al., 1980). Thus, hyperventi-
11.4. Slow and shallow breathing lation appears to signify an unsuccessful outcome
of the coping process. This type of coping is often
A decrease in RR and depth has been associ- called passive coping because the individual
ated with emotions like passive grief and calm has no other choice than to try to accommodate
happiness, in which physical exertion is aban- to the stressor, or to try to reduce its impact as
doned (Averill, 1969). This finding is supported much as possible (Obrist, 1981). Passive coping
by a number of studies. Decrease in RR and may often involve strong (but inhibited) action-
depth of breathing is found in depression, during oriented tendencies, and may be accompanied by
calm pleasurable experiences and during states of anxiety, fear, anger or panic. It has often been
relaxation (Rehwoldt, 1911; Nakamura, 1984). On suggested that the hyperventilatory response may
the other hand, clinical chronic depression has functionally be interpreted in terms of prepara-
been found to be related with an increase in RR tion for fight or flight (Suess et al., 1980; Dudley,
and decreased levels of pC0, (Damas-Mora et 1964). This type of response by itself seems in-
al., 1976, 1982). However, as we noted before, deed to be physiologically functional if fight or
this apparent paradox might be due to the anxiety flight would ensue (by way of anticipating en-
component that is associated with clinical depres- hanced 0, consumption and CO, production).
sion. Thus, slow and shallow breathing and de- However, when physical action is blocked or sus-
creased respiratory drive appears to be primarily pended, the anticipated increase in 0, consump-
indicative of states that are characterized by with- tion and CO, production does not materialize.
drawal from the environment and by passiveness, As a consequence, the ventilatory response is in
and might be encountered during depressed and excess of metabolic needs and may be therefore
unhappy, as well as happy, but unexcited moods. be regarded as non-functional

11.5. Normocentilation and hyperL)entilation Il. 6. Thoracic-abdominal balance

Breathing patterns may not only be character- The psychological correlate of thoracic-ab-
ized in terms of variations in rate and depth, but dominal balance might either be hedonic tone
also, on another level of description, by the de- (variations along a pleasantness-unpleasantness
gree to which ventilation is in accordance with dimension), or inhibitory control. Thoracic domi-
metabolic requirements, i.e. with CO, produc- nance appears primarily to correspond to un-
tion. In this regard, there is an important differ- pleasant affect, tenseness or anxiety, and abdomi-
ence between normoventilatory and hyperventila- nal dominance to pleasant emotional states or
tory responses. Normoventilatory responses relaxation (Ancoli and Kamiya, 1979; Ancoli,
(which are identified by stable end-tidal pC0, Kamiya and Ekman, 1980; Lum, 1976; Svebak,
levels that remain within the normal range) seem 1975).
Il. 7. Irregularity of breathing cal difficulties that may be hard to overcome and
that, surely, have not been overcome in the stud-
A number of studies has shown that breathing ies reviewed. The major obstacles that are com-
may become irregular under conditions of emo- monly encountered when investigating emotion-
tional upset, excitement, and task involvement respiration relationships are due to problems with
(e.g., Hormbrey et al., 1986; Santibafiez and regard to intensity scaling and dimensional inter-
Bloch, 1986; Stevenson and Ripley, 1952; Tobin pretation. The intensity scaling problem refers to
et al., 1988; Boiten, 1993b). Marked breath-to- the fact that observed differences between cmo-
breath variations in Vt, RR and Vt/Ti are quite tional conditions may only reflect differences in
common during normal breathing (Bradley, 1977; intensity. Emotions characteristically differ in av-
Milic-Emili et al., 1981; Tobin et al., 1988). It erage intensities and in intensity ranges (Frijda ct
appears that variation in respiratory parameters al., 1992). Typical fear, for example, tends to be
is not an entirely random process since oscillatory more intense than typical anger (Scherer et al.,
patterns in the breath to breath variability of 1986).
respiration have been observed in healthy sub- The problem with regard to dimensional inter-
jects at rest (Lenfant, 1967). Oscillatory or cycli- pretation resolves around the fact that responses
cal changes in respiratory parameters have been to particular emotional conditions may solely re-
related to respiratory feedback mechanisms such flect variations in the location of the resulting
as peripheral and central chemoreflex activity emotional states on dimensions like negative or
(van den Aardweg, 1992). However, very little is positive affect, excitement, tenseness, motor
known about the mechanisms that augment the preparation, or readiness for active coping. As we
occurrence of respiratory irregularities during indicated earlier, these locations may not charac-
emotional behaviour. It might be speculated that terize particular emotions as such, but only the
enhanced breathing irregularities ensue when accidental exemplars of fear, anger, etc. that were
neural influences upon respiratory control pro- included in the particular study. Indeed, consid-
cesses are highly incompatible, as may be the case ering the broad spectrum of emotional stimuli
during emotional and behavioural states that in- utilized in, for instance, the studies involving anx-
volve conflicting, or incongruous components. iety, it is not surprising that the results are often
inconsistent. Anxiety has been evoked by threat-
ening subjects with electric stimulation, by admin-
11.8. Emotion differentiation istering electric stimulation, by hypnotic sugges-
tion, films, and among clinical conditions. The
We reviewed the weak evidence, mostly dating types and manifestations of evoked anxiety may
from the beginning of this century, that emotions well be different in many relevant aspects. An
can be distinguished and identified by specific illustration is provided by Dudley et al. (1964).
patterns of respiratory parameters. However, the The induction of intense pressure headache pro-
only solid case of a specific emotion-respiration duced respiratory hyperfunction only when it
connection that we encountered is that of the elicited overt, action oriented anger or anxiety.
breathing suspension in startle or surprise. The To truly establish relationships between particu-
notion that specific patterns of respiratory (and lar emotions and respiratory (or more generally,
other physiological) variables might be character- physiological) patterns, one needs an experimen-
istic for different emotions is, of course, attractive tal design that controls for the effects of specific
and plausible, and consonant with the James- stimulus situations and specific response require-
Lange theory of emotion. However, the scant ments. As it was phrased by Averill (1969, p. 411):
evidence allows no firm conclusion that the no- A superior approach to the problem of differen-
tion is a valid or invalid one. Studies comparing tial physiological patterning during emotion would
the psychophysiological response patterns of be to use an experimental design in which each
emotions are plagued by conceptual and empiri- emotion is compared with itself under different
 F.A. Boiten et al. /International Jo~mul of Psychophysiology 17 (1994) 103-128 123

stimulus conditions, as well as with other emo- former type of respiration appears generally tied
tions. to adaptive or active coping behaviour, and the
latter to be typical for passive or unsuccessful
coping. Thus, taking all the available evidence
12. Conclusions together, the patterning of respiratory responses
may hold potential as a diagnostic tool to distin-
Although the literature that was reviewed ob- guish a number of important dimensions of varia-
viously suffers from a number of important limi- tion of emotional response.
tations, the available evidence on emotion-respi- With regard to the two-dimensional models of
ration relationships contains several elements of emotion, those of Russell (1980) and of Watson
general interest. One of the most important con- and Tellegen (198.51, the evidence from respira-
clusions that can be drawn, is that certain pat- tion favors the former rather than the latter.
terns of respiratory responses appear to be corre- Respiratory changes correspond to hedonic tone
lated with dimensions that define general aspects and level of activation, rather than to positive
of emotions. affect and to negative affect. This conclusion is of
As a first approximation, we noted marked interest, because also on other grounds it can be
differences between patterns suggestive of respi- argued that, while Watson and Tellegens rota-
ratory hypo- and hyperfunction that appear to be tion of the bidimensional plot appears to reflect
tied to specific instances of passive/ depressed, the ecology of emotional states, Russels (1980)
and active/excited affect states (whereby relaxed rotation better reflects basic emotional processes
emotional states seem to correspond to respira- (Frijda, 1993).
tory responses that may be located in the middle Although, as we noted above, the specificity
region of the continuum). In a different formula- model of respiration-emotion relationships can-
tion, this range of affect states extends from not be rejected, the available evidence is more
passive withdrawal from the environment to readily compatible with a dimensional or re-
active engagement. sponse requirement view. The evidence is most
Within the realm of respiratory hyperfunction easily interpreted in terms of functional response
(which is typified by a high mean inspiratory flow modes (excitement, active coping etc.); strong
rate), we have furthermore found that there may similarities exist between the respiratory patterns
be a continuum of respiratory responses extend- in cognitively oriented tasks and tasks designed to
ing from rapid shallow breathing to rapid deep elicited specific emotions, as well as between
breathing. Emotionally, this continuum appears those in widely diverging emotions. Against the
to represent a range of affects that extends from dimensional model, in the way we defined it
inhibition, controlled activation and tenseness on earlier, and to the extent that it is different from
the one extreme, to excitement, unstrained acti- a response requirement model, one may hold the
vation and overt panic on the other. Inhibition fact that no pattern of respiratory response, so
and activation, of course, are important facets of far, has distinguished between pleasant and un-
behavioural energetics, existing next to atten- pleasant states.
tional or sympathetic arousal, and independent of The emotion aspects found to be reflected in
it (Pribram, 1981). The respiratory dimension may respiration are fairly crude. Further differentia-
be of general interest, in that diagnostic measures tion, either between emotions or between aspects
of inhibition or tenseness, and of unrestrained of behavior energetics, probably requires inclu-
activation, are not otherwise readily available. sion of additional respiratory parameters to those
Lastly, there appears to be still another impor- usually measured, and application of sophisti-
tant dimension of respiration, distinguishing be- cated analysis techniques. As we have seen, tradi-
tween breathing that is metabolically appropriate tional respiratory analysis techniques are limited
(normoventilation), and breathing that is in ex- to assessment of changes in RR, depth, and Vt.
cess of metabolic demands (hyperventilation). The The main disadvantage of these techniques is that
124 F.A. Boiten et al. /International Journal of Psychophysiology 17 (1994) 103-128

they do not provide insight in the mechanisms cal differentiation of positive and negative affects.
Biofeedback Self-Regulation, 5: 356-357 (Abstract).
which determine Vt and RR (Milic-Emili et al.,
Askanazi, J., Silverberg, P.A., Forster. R.J.. Hyman, A.I.,
1981). Therefore, analysis of respiratory parame- Milic-Emili, J. and Kinney, J.M. (19801 Effects of respira-
ters should proceed along the lines that were tory apparatus on breathing pattern. J. Appl. Physiol., 48:
proposed by Milic-Emili and his co-workers (1976, 577-580.
19811, and that we have described in the section Averill, J.R. (1969) Autonomic response patterns during sad-
ness and mirth. Psychophysiology, 5: 399-414.
on respiratory parameters. Moreover, rate, depth
Ax, A.F. (1953) The physiological differentiation between fear
and volume may not be sufficient nor subtle
and anger in humans. Psychosom. Med., 15: 433-442.
enough to detect emotion-relevant differences. Bachy, J.P.A., Baconnier, P. and Benchetrit, G. (1986) A
Additional phasic components of the breathing program for cycle-by-cycle analysis of biological rhythms.
cycle such as saccadic respiratory movements, Application to respiratory rhythm. Comput. Methods Pro-
gram. Biomed., 23: 297-307.
hypopnoea, apnoea, inspiratory and expiratory
Baker, D.M. (1934) Sighing respiration as a symptom. Lancet,
pauses, the inspiration-expiration angle, and sigh- 1: 174-179.
ing might have to be included in the analyses. Bartlett, R.G. (1956) Physiologic responses during coitus. J.
The effects of inhibitory processes could probably Appl. Physiol., 9: 469-472.
be evaluated by morphological parameters such Barry, R.J. (1982) Novelty and significance effects in the
fractionation of phasic OR measures: a synthesis with
as angularity of the respiratory curve (Clausen,
traditional OR theory. Psychophysiology, 19: 28-35.
1951; Gautier, 19801, and disturbances of control Bass, C. and Gardner, W.N. (1985) Emotional influences on
processes by irregularities in breathing. With the breathing and breathlessness. J. Psychosom. Res.. 29: 592%
aid of modern computer procedures and tech- 609.
niques, reliable component analyses of the Bell, Ch. (1806) The anatomy and philosophy of expression,
London, Bohn.
breathing cycle can be undertaken (see Cohen et
Benchetrit, G., Shea, S.A., Pham Dinh, T., Bodocco, S..
al., 1975; Wientjes et al., 1988; Boiten, 1993a,b). Baconnier, P. and Guz, A. (1989) Individuality of breath-
Investigating these additional parameters is both ing patterns in adults assessed over time. Respir. Physiol.,
desirable and feasible. 75: 199-210.
Bendixen, H.H., Smith, G.M. and Mead, M. (1964) Pattern of
ventilation in young adults. J. Appl. Physiol., 19: 195510X.
Benussi, V. (1914) Die Atmungssymptome der Liige. Arch.
ges. Psychol., 31: 244-273: 396-419.
References
Blatz, W.E. (1925) The cardiac respiratory, and electrical
phenomena involved in the emotion of fear. J. Exp. Psy-
Adamson, J.D., Romano, K.R., Burdick, J.A., Corman, CL. chol., 8: 109-132.
and Chebib, F.S. (1972) Physiological responses to sexual Bloch, S.. Lemeignan, M. and Aguilerat, N. (1991) Specific
and unpleasant film stimuli. J. Psychosom. Res., 16: 153- respiratory patterns distinguish among human basic emo-
162. tions. Int. J. Psychophysiol., 11: 141-154.
Alechsieff, N. (1907) Die Grundformen der Gefiihle. Psycho]. Boiten, F.A. (1993a) Component analysis of task related respi-
Stud., 3: 1566271. ratory patterns. Int. J. Psychophysiol., 15: 91-104.
Alexander, F. and Saul, L.J. (1940) Respiration and personal- Boiten, F.A. (lY93b) Emotional breathing patterns. Doctoral
ity: a preliminary report: part I. Description of the curves. Dissertation, Amsterdam.
Psychosom. Med., 2: 110-118. Boiten, F.A. (submitted) Autonomic response patterns during
Allen, M.T., Sherwood, A. and Obrist, P.A. (1986) Interac- voluntary facial action. Psychophysiology.
tions of respiratory and cardiovascular adjustments to be- Bradley, G.W. (1977) Control of the breathing pattern. In J.G.
havioral stressors. Psychophysiology, 23: 532-541. Widdicombe (Ed.), MTP Int. Review of Science: Ser. I,
Allen, M.T. and Crowell, M.D. (1989) Patterns of autonomic Physiology; Vol. 2, Respiratory Physiology II, University
response during laboratory stressors. Psychophysiology, 26: Park Press, Baltimore, pp. 1855217.
603-614. Buikhuysen, M. and Garssen, B. (1990) Hyperventilation and
American Psychiatric Association (1980) Diagnostic and sta- panic attacks. Biol. Psychol., 31: 280. (Abstract).
tistical manual of mental disorders. APA, Washington, Burns, B.H. (1971) Breathless in depression. Brit. J. Psychiat.,
DC. 119: 39-45.
Ancoli, S. and Kamiya, J. (1979) Respiratory patterns during Cabot, R.C. and Adams, F.D. (1942) Physical Diagnosis, 13th
emotional expression. Biofeedback Self-regulation, 4: 242 edn., Baltimore, Williams and Wilkins.
(Abstract). Cannon, W.B. (1929) Bodily changes in pain, fear and rage,
Ancoli, S., Kamiya, J. and Ekman, P. (1980) Psychophysiologi- 2nd edn., New York, Appleton and Company.
 F.A. Boiten et al. /International Journal of Psychophysiology I7 (1994) 103-128 125

Carroll, D, Turner, J.R. and Hellawell, J.C. (19861 Heart rate Drozynski, L. (19111 Atmungs- und Pulssymptome rytmischer
and oxygen consumption during active psychological chal- Gefihle. Psychol. Studien, 7: 83-140.
lenge: the effects of level of difficulty. Psychophysiology, Dudley, D.L., Holmes, T.H., Martin, C.J. and Ripley, H.S.
23: 174-181. (19641 Changes in respiration associated with hypnotically
Carroll, D.. Turner, J.R. and Rogers, S. (1987) Heart rate and induced emotion, pain and exercise. Psychosom. Med., 26:
oxygen consumption during mental arithmetic, a video 46-57.
game. and graded static exercise. Psychophysiology, 24: Dudley, D.L. (1969) Psychophysiology of respiration in health
112-118. and disease, New York, Appleton-Century-Crofts.
Chadha. T.S., Watson, H., Birch, S., Jenouri. G.A., Schneider, Dumas, G. (1948) La vie affective, Paris, Presses Universi-
A.W., Cohn, M.A. and Sackner, M.A. (1982) Validation of taires de France.
respiratory inductive plethysmography using different cali- Euler, C. von (1977) The functional organization of the respi-
bration procedures. Amer. Rev. Respir. Dis., 125: 644-649. ratory phase-switching mechanisms. Fed. Proc., 36: 2375-
Christie, R. (1935) Some types of respiration in the neuroses. 2380.
Quart. J. Med., 4: 427-432. Faulkner, W.B. (19411 The effect of the emotions upon di-
Clark, D.M., Salkovskis, P.M. and Chalkley, A.J. (19851 Res- aphragmatic function. Psychosom. Med., 3: 187-189.
piratory control as a treatment for panic attacks. J. Behav. Feleky, A. (1914) The influence of the emotions on respira-
Ther. Exp. Psychiat., 16: 23-30. tion. Psychol. Rev., 1: 218-241.
Clark, D.M., Salkovskis, P.M. Gelder, M., Koehler, C.. Mar- Fenz, W.D. and Jones, G.B. (19721 Individual differences in
tin, M., Anastasiades, P., Hackman, A.. Middleton, H. and physiological arousal and performance in sport
Jeavons, A. (1988) Tests of a cognitive theory of panic. In parachutists. Psychosom. Med., 34: l-8.
I. Hand and H.U. Wittchen (Eds.1, Panic and Phobias II, Finesinger, J.E. (1939) The effects of pleasant and unpleasant
Springer-Verlag, New York. ideas on the respiratory pattern (spirogram) in psychoneu-
Clausen, J. (1951) Respiration movement in normal, neurotic rotic patients. Arch. Neurol. Psychiat., 42: 659-667.
and psychotic subjects. Acta Psychiat. Neurol., Suppl. 68. Finesinger, J.E. (1943) The spirogram in certain psychiatric
Cohen, H.D., Goodenough, D.R., Witkin, H.A., Oltman, P., disorders. Amer. J. Psychiat., 100: 159-169.
Gould, H. and Shulman, E. (1975) The effects of stress on Finesinger, J.E. and Mazick, SC. (19401 The respiratory
components of the respiration cycle. Psychophysiology, 12: response of psychoneurotic patients to ideational and to
3777380. sensory stimuli. Amer. J. Psychiat., 97: 27-46.
Cohn, M.A., Rao, A.S.V., Broudy, M., Birch, S., Watson, H., Freeman, L.J., Conway, A. and Nixon, P.G.F. (1986) Physio-
Atkins, N., Davis, B., Stott, F.D. and Sackner, M.A. (19821 logical responses to psychological challenge under hypno-
The respiratory inductive plethysmograph: A new non sis in patients considered to have the hyperventilation
invasive monitor of respiration. Bull. Eur. Physiopathol. syndrome: implications for diagnosis and therapy. J.R.
Respirat., 18: 643-658. Sot. Med., 79: 76-83.
Cooper, K.E., Martin, S. and Riben, P. (19761 Respiratory Friedman, M. (1945) Studies concerning the etiology and
and other responses in subjects immersed in cold water. J. pathogenesis of neurocirculatory asthenia. IV. The respi-
Appl. Physiol., 40: 903-Y 10. ratory manifestation of neurocirculary asthenia. Amer.
Coppen, A.J. and Mezey, A.G. (19601 The influence of sodium Heart J., 30: 557-566.
amytal on the respiratory abnormalities of anxious psychi- Frijda, N.H. (1986) The emotions, Cambridge, Cambridge
atric patient. J. Psychosom. Res., 5: 52-55. University Press.
Corwin, W. and Barry, H. (1940) Respiratory plateaux in Frijda, N.H. (1993) Moods, emotion, episodes, and emotions,
day-dreaming and in schizophrenia. Amer. J. Psychiat., In J. Haviland and M. Lewis (Eds.), Handbook of emo-
97: 308-318. tions, New York, Guilford Press, pp. 381-403.
Craig, K.D. (1968) Physiological arousal as a function of Frijda, N.H., Kuipers, P. and Terschure, El. (1989) Relations
imagined, vicarious, and direct stress experiences. J. Ab- between emotion, appraisal, and emotional action readi-
norm. Psychol., 73: 513-520. ness. J. Pers. Sot. Psychol.. 57: 212-228.
Damas-Mora, J., Grant, L., Kenyon, P., Patel, M.K. and Frijda, N.H., Ortony, A., Sonnemans, J. and Clore, G. (19921
Jenner, F.A. (19761 Respiratory ventilation and carbon The complexity of intensity. In M. Clark (Ed.), Emotion.
dioxide levels in syndromes of depression. Brit. J. Psychiat., Review of personality and social psychology, Vol. 13, Bev-
129: 457-469. erley-Hills, Sage, pp. 60-89.
Damas-Mora, J., Souster, L. and Jenner, F.A. (1982) Dimin- Gardner, W.N., Meah, M.S. and Bass, C. (1986) Controlled
ished hypercapnic drive in endogenous or severe depres- study of respiratory responses during prolonged measure-
sion. J. Psychosom. Res., 26: 237-245. ment in patients with chronic hyperventilation. Lancet, ii:
Darwin, Ch. (1892) The expression of the emotions in man 826-830.
and animals, London, John Murray (1965, Chicago, Uni- Garssen, B. (1986) Psychofysiologie van de ademhaling en het
versity of Chicago Press) Hyperventilatiesyndroom (Dutch), Doctoral Thesis, Uni-
Douglas, N.J., White, D.P., Picket& C.K., Weil, HJ.V. and versity of Utrecht.
Zwillich, L.W. (1982) Respiration during sleep in normal Gautier, H. (1980) Control of the pattern of breathing. Clin.
man. Thorax, 37: 840-844. Sci., 5X: 343-348.
Gilbert. B., Auchincloss, J.H.. Brodsky, J. and Boden, W.. Koegler. R.R. and Kline, L.Y. (1065) Psychotherapy research:
1972. Changes in tidal volume, frequency and ventilation an approach utilizing autonomic response measures. Amer.
induces by their measurement. J. Appl. Physiol.. 33: 252- J. Psychother., 19: 26X-279.
254. Lacey, J.1. and Lacey, B.C. (1970) Some autonomic-central
Goldstein. M.J.. Jones. R.B., Clemens, T.L.. Flagg, G.W. and nervous system relationships. In P. Black (Ed.), Physiologi-
Alexander, F.G. (1965) Coping style as a factor in psy- cal correlates of emotion, New York, Academic Press. pp.
chophysiological response to a tension-arousing film. J. 205-228.
Pers. Sot. Psycho].. I: 290-302. L.andia, C. (1926) Studies of emotional reactions. V. Severe
Greenfield, N.S. and Sternbach, R.A (lY72) Handbook of emotional upset. J. Comp. Paychol., 6: 221-242.
Psychophysiology, New York, Holt, Rinehart and Winston. L.ane, J.D. (1981) The effects of self-control of respiration
Gribbin, H.R. (lYX3) Using body surface movements to study upon experimental pain. Psychophysiology. IX: 162-163
breathing. J. Med. Engineering and Technology, 5: 217- (Abstract).
223. Lang. P.J.. Bradley, M.M. and Cuthbert, B.N. (IYYO) Emotion.
Grossman. P. (1983) Respiration, stress and cardiovascular attention, and the startle reflex. Psycho]. Rev., Y7: 377-395.
function. Psychophysiology. 20: 2844300. Langer, A.W.. McCubbin, J.A., Stoney, C.M., Ilutcheson. J.S.,
Grossman. P., de Swart. J.C.G. and Defares, P.B. (1985) A Charlton, J.D. and Obrist, P.A. (1985) Cardiopulmonary
controlled study of breathing therapy for treatment of adjustments during exercise and an aversive reaction time
hyperventilation syndrome. J. Psychosom. Res.. 2Y: 4Y-5X. task: effects of beta-adrenoceptor blockade. Psychophysi-
Grossman, P. and Wientjes, C.J.E. (198)) Respiratory disor- ology, 22: 59-68.
dcra: asthma and hyperventilation syndrome. In G. Turpin Lehmann, A. tlYl4) Die fIauptgesetze des menschlichen
(Ed.), Handbook of Clinical Psychophysiology, Chichester, Gefiihlslebens, Leipzig, Reisland.
Wiley, pp. 519-554. Lehman, H.E. (lY79) Yawning a homeostatic reflex and its
Ilarrer, G. (1969) Atmung und Psyche. Wiener Med. Wochen- psychological significance. Bull. Menniger Clinic. 43: l23-
schrift, 5 I /S2: X84-887. 136.
Harver, A. and Kotses, H. (1987) Effects of auditory stimula- Lenfant, C. (1967) Time-dependent variations of pulmonary
tion on respiration. Psychophysiology, 24: 2634. gas exchange in normal man at rest. J. Appl. Physiol., 22:
Hogan, D.H. (1970) Some physiological determinants of auto- 675-684.
nomic responsivity to sound. J. Speech and Hearing Res.. Levenson. R.W.. Ekman. P. and Friesen, W.V. (1990) Volun-
13: 130-146. tary facial action generates emotion-specific autonomic
Holmes, D.S., McCaul, K.D. and Solomon, S. (1978) Control nervous system activity. Psychophysiology, 27: 363-384.
of respiration as a means of controlling responses to Ley, R. (1985) Agoraphobia, the panic attack, and the hyper-
threat. J. Pers. Sot. Psycho]., 36: 19X-204. ventilation syndrome. Behav. Res. Ther., 23: 29-31.
Hormbrey. J.M., Patil. C.P.. Jacobi, MS. and Saunders, K.B. Lum. L.C. (1976) The syndrome of habitual hyperventilation.
(1976) Hyperventilation Syndrome Analysis of irregular- In O.W. Hill (Ed.). Modern trends in Psychosomatic
ity of breathing. Bull. Eur. Physiopathol. Respirat., 22: medicine, London, Butterworth.
518. Lynch, J.J., Thomas, S.A., Long, J.M., Malinow. K.L.. Chick-
Hornsveld. H., Garssen, B.. Fiedeldy Dop, M. and Spiegel, P. adonz, G. and Katcher. A.II. (19X0) Human speech and
van. (1990) Symptom reporting during voluntary hyperven- blood pressure. J. Nerv. Ment. Dis., 16X: 526534.
tilation and mental load: implications for diagnosing hy- Mador. M., J. and Tobin, M.J. (1YYl) Effects of alterations in
perventilation syndrome. J. Psychosom. Res., 34: 6X7-697. mental activity on the breathing pattern in healthy sub-
James, W. (1884) What is an emotion? Mind, 9: 18X-205. jects Amer. Rev. Respir. Dis.. 144: 481-4X7.
Jones. M.S. and Scarisbrick, R. (1941) A comparison of the Malmo. R.B. and Shagass, C. (1940) Physiologic studies of
respiration in patients with effort syndrome and in normal reaction to stress in anxiety and early schizophrenia. Psy-
subjects. Proc. R. Sot. Med., 34: 54YY553. chosom. Med., I I: 9-24.
Kagan, J. and Rosman. B.L. (1964) Cardiac and respiratory Masters, W. and Johnson, V. (1966) Human sexual response.
correlates of attention and an analytic attitude. J. Exp. Boston, Little, Brown and Co.
Child Psychol., 1: 50-63. Mayer, C. (I921 J Physiologisches und Pathologisches Llber
Kay, J.D.S., Strange Petersen, E. and Vejby-Christensen, H. das Gahnen. Zeitsch. Biol.. 73: 101-l 14.
(1975) Breathing in man during steady-state exercise on Mekjavic, I.B., La Prairie, A., Burke, W. and Lindborg, B.
the bicycle at two pedalling frequencies, and during tread- (1987) Respiratory drive during sudden cold water immer-
mill walking. J. Physiol., 251: 645-656. sion. Respir. Physiol., 70: 121~130.
Kay. J.D.S., Strange Petersen, E. and Vejby-Christensen, H. Milic-Emili. J., Grassino, A.E. and Whitelaw, W.A. (1081)
(lY75) Mean and breath-by-breath pattern of breathing in Measurement and testing of respiratory drive. In T.F.
man during steady-state exercise. J. Physiol., 251: 6577669. Hornbein (Ed.), Regulation of Breathing, Part II. New
Keatinge, W.R. and Nadel, J.A. (1965) Immediate respiratory York, Dekker, pp. 675-744.
response to sudden cooling of the skin. J. Appl. Physiol., Milic-Emili, J. and Grunstein, M.M. (lY76) Drive and timing
20: 65-6Y. components of ventilation. Chest, 70: 131-133.
 F.A. Boiten et 01. /International Joumul of Psychophysiology I7 (1994) 103-12X 127

Morgan, W.P. (lY83) Psychometric correlates of respiration: a Scherer, K.R.. Walbott, H.G. and Summerfield, A.B. (1986)
review. Amer. Industrial Hygienic Assoc. J., 44: 677-6X4. Experiencing emotions: A cross-cultural study, Cambridge.
Morgan. W.P. (1985) Psychogenic factors and exercise Cambridge University Press.
metabolism: a review. Med. Sci. Sports Exercise, 17: 309- Schnore, M.M. (1950) Individual patterns of physiological
316. activity as a function of task differences and degree of
Morel. D.R., Forster. A. and Suter, P.M. (1983) Noninvasive arousal. J. Exp. Psychol.. 5X: 117-128.
ventilatory monitoring with bellows pneumographs in Shershaw. J.C.. Janarek, D.J. and Kazemi, H. (1976) Ventila-
supine subjects. J. Appl. Physiol., 55: 598-606. tory response to carbon dioxide inhalation in depression.
Nakamura, H. (19841 Effects of musical emotionality upon Psychosom. Med., 3X: 2X2-287.
GSR and respiration rate: the relationship between verbal Shershaw, J.C.. King, A. and Robinson, S. (lY73) Carbon
reports and physiological responses. Shinrigaku Kenkyu, dioxide sensitivity and personality. Psychosom. Med., 35:
55: 47-50. 1555160.
Neiss, R. (1988) Reconceptualizing arousal: Psychobiological Sherwood, A.. Allen, M.T.. Obrist, P.A. and Langer, A.W.
states in motor performance. Psychol. Bull., 103: 345366. (1986) Evaluation of beta-adrenergic influences on cardio-
Neiss. R. (1990) Ending arousals reign of terror: A reply to vascular and metabolic adjustments to physical and psy-
Anderson. Psychol. Bull., 107: 101-105. chological stress. Psychophysiology, 23: X9-104.
Newson-Davis, J. and Stagg, D. (197.5) Interrelationships of Sims. J., Carroll, D.. Turner, J.R. and Hewitt. J.K. (19X8)
the volume and time components of individual breaths in Cardiac and metabolic activity in mild hypertensive and
resting man. .I. Physiology (Land.), 245: 481-498. normotensive subjects. Psychophysiology. 25: 172-17X.
Obrist, P.A. (1981) Cardiovascular psychophysiology: A per- Skaggs. E.B. (1930) Studies in attention and emotion. J.
spective. New York, Plenum Press. Comp. Psychol., IO: 375-419.
Perez, W. and Tobin, M.J. (1985) Separation of factors re- Skarbek. A. (19701 A psychophysiological study of breathing
sponsible for change in breathing pattern induced by in- behaviour. Brit. J. Psychiat., 116: 637-641.
strumentation. J. Appl. Physiol., 5Y: 1515-1520. Snyder. F. and Scott. J. (1072) The psychophysiology of sleep.
Porges. S.W. and Raskin, D.C. (1969) Respiratory and heart In N.S. Greenfield and R.A. Sternbach tEds.1, Handbook
rate components of attention. J. Exp. Psychol., Xl: 4977503. of Psychophysiology. New York, Holt. Rinehart and Win-
Pribram, K.H. (1981) Emotions. In S.B.Filskov and T.J. Boll ston. pp. 645708.
(Eds.1, Handbook of clinical neuropsychology. New York, Sokolov, J.N. (1963) Perception and the conditioned reflex.
Wiley. pp. 1022134. Oxford, Pergamon Press.
Provine. R.R. and Hamernik. H.B. (1986) Yawning: effects of Spielberger, C.D., Johnson. E.H.. Russell. S.F., Crane, R.J..
stimulus interest. Bull. Psychonomic Sot., 24: 437-438. Jacobs, G.A. and Worden. T.J. (1985) The experience and
Rehwoldt, F. (IYl I) Uber respiratorische Affektsymptome. expression of anger: Construction and validation of an
Psychol. Studien 7: 141-195. anger expression scale. In M.A. Chesney and R.H. Roscn-
Reich. W. (1949) Character analysis, New York. Orgone Insti- man (Eds.), Anger and hostility in cardiovascular and
tute Press. behavioral disorders, New York, Hemisphere/McGraw-
Rime. B., Philippot. P. and Cisamolo, D. tlYY0) Social Hill, pp. S-30.
schemata of peripheral changes in emotion. J. Pers. Sot. Stein, M.D. and Luparello, M.D. (1967) The measurement of
Psychol.. 59: 3X-49. respiration. In C. Clinton and Ph. D. Brown (Eds.). Meth-
Rimm. D.C. and Litvak, S.B. (1969) Self-verbalization and ods in Psychophysiology, Baltimore. Williams and Wilkins.
emotional arousal. J. Abnorm. Psychol.. 74: 1X1-187. pp. 75-94.
Ruckmick, C.A. (19361 The psychology of feeling and emo- Stevenson, I. and Ripley. H.S. (1952) Variations in respiration
tion, New York. McGrawHill. and in respiratory symptoms during changes in emotion.
Russell, J.A. 1980. A circumplex model of affect. J. Pers. Sot. Psychosom. Med., 14: 476-490.
Psychol., 39: 1161-117X. Stoerring, G. (lYO6) Experimentelle Beitrage zur Lehre vom
Sackner, M.A. (IYXOI Monitoring of ventilation without physi- Gefiihl. Arch. ges. Psychol.,h: 316-356.
cal connection to the airway: a review. In F.D. Scott (Ed.), Suess, W.M.. Alexander, A.B., Smith, D.D., Sweeney, H.W.
ISAM Proceedings of the Second International Sympo- and Marion, R.J. (1980) The effects of Psychological stress
sium on Ambulatory Monitoring. London, Academic Press. on respiration: A preliminary study of anxiety and hyper-
pp. 29993 IO. ventilation. Psychophysiology, 17: 5355540.
Santibancz-H, G. and Bloch, S. (19861 A qualitative analysis Suter. J. (1912) Die Beziehungen zwischen Aufmcrksamkeit
of emotional effector patterns and their feedback. Pavlov. und Atmung. Arch. ges. Psychol.. 25: 7X-150.
J. Biol. Sci.. 21: 10X-116. Svebak, S. (1975) Respiratory patterns as predictors of laugh-
Schachter, J. (1957) Pain, fear, and anger in hypertensives and ter. Psychophysiology. 12: 62-65.
normotensives. Psychosom. Med., 19: 17-29. Svebak, S. (1982) The effects of difficulty and threat of
Scherer, K.R. (1986) Vocal affect expression: A review and a aversive electric shock upon tonic physiological changes.
model for future research. Psychol. Rev.. 9Y: 1433165. Biol. Psychol., 14: 113-128.
128 F.A. Boiten et al. /International Journal of Psychophysiology 17 (1994) 103-128

Timmons, B., Salamy. .I.. Kamiya. J. and Girton, D. (1972) Watson, D., Clark, L.A. and Carey. G. (1988) Positive and
Abdominal-thoracic respiratory movements and levels of negative affectivity and their relation to anxiety and de-
arousal. Psychon. Sci., 27: 1733175. pressive disorders. J. Abnorm. Psychol., 97: 3466353.
Tobin, M.J., Chadha, T.S., Jenouri, G., Birch, S.J., Gazeroglu, Watson, D. and Pennebaker, J.W. (1989) Health complaints.
H.B. and Sackner, M.A. (1983) Breathing patterns. 2. stress and distress: Exploring the central role of negative
Diseased subjects, Chest, 84: 2X6-294. affectivity. Psychol. Rev., 96: 234-254.
Tobin, M.J., Mador, M.J., Guenther, S.M., Lodato, R.F. and Watson, D. and Tellegen, A. (1985) Towards a consensual
Sackner, M.A. (1988) Variability of resting respiratory structure of mood. Psychol. Bull., 9X: 219-235.
drive and timing in healthy subjects. J. Appl. Physiol., 65: Western, P.J. and Patrick, J.M. (1987) Effects of focusing
3099317. attention on breathing with and without apparatus on the
Tucker, W.D. (1963) Hyperventilation in differential diagno- face. Respir. Physiol., 72: 123-130.
sis Med.Clinics, 47: 491-497. Wientjes, C.J.E. (1992) Respiration in psychophysiology:
Turner, J.R., Carroll, D. and Courtney, H. (1983) Cardiac and Methods and applications. Biol. Psychol., 34: 179-203.
metabolic responses to space invaders: an instance of Wientjes, C.J.E. (1993) Psychological influences upon breath-
metabolically-exaggerated cardiac adjustment. Psychophys- ing: Situational and dispositional aspects, Doctoral Disser-
iology, 20: 544-549. tation, Tilburg.
Turner, J.R. and Carroll, D. (1985) Heart rate and oxygen Wientjes, C.J.E., Grossman, P. and Meijden, P. van der (IYXX)
consumption during mental arithmetic, a video game. and A computer program for breath-by-breath analysis of car-
graded exercise: further evidence of metabolically-exag- diorespiratory parameters. IZF Report 1988 1-2, TN0
gerated cardiac adjustment. Psychophysiology, 22: 261-267. Institute for Perception, Soesterberg, the Netherlands.
Turpin. G. (1986) Effects of stimulus intensity on autonomic Wilier, J.C. (1975) The influence of anticipation of pain on
responding: the problem of differentiating orienting and heart and respiratory rates and on the pain reflex in man.
defense reflexes. Psychophysiology, 23: l-14. Physiol. Behav., 15: 411-415.
Van den Aardweg, J.G. (1992) Cardiovascular effects of cyclic Woodworth. R.S. (1938) Experimental Psychology, New York,
changes in breathing pattern, Doctoral Thesis, University Holt.
of Amsterdam. Woodworth, R.S. and Schlosberg, H. (lY55) Experimental
Venables, P.H. (1984) Arousal: An examination of its status as Psychology (Revised Ed.), New York, Holt.
a concept. In M.G.H. Coles, J.R. Jennings and J.A. Stern Wundt, W. (1902) Grundziige der physiologischen Psycholo-
(Eds.), Psychological perspective: Festschrift for Beatrice gie, Vol. 3 (5. Ausgabe), Leipzig, Engelmann.
and Jonn Lacey. New York, Van Nostrand Reinhold. pp. Zuckerman, M. (1971) Physiological measurements of sexual
1344142. arousal in the human. Psychol. Bull., 75: 297-32).
Watson. H. (1980) The technology of inductive plethysmogra- Zuckerman, M., Persky, H. and Curtis, G.C. (lY68) Relation-
phy. In F.D. Stott (Ed.), ISAM Proceedings of the Second ships among anxiety, depression, hostility and autonomic
International Symposium on Ambulatory Monitoring. variables. J. Nervous Mental Dis.. 140: 481-4X7.
London, Academic Press.