End-tidal versus transcutaneous measurement of PCO
2
during voluntary
hypo- and hyperventilation
Eileen Wollburg
a,b,c,
⁎
, Walton T. Roth
a,b
, Sunyoung Kim
a,b
a
Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, CA, USA
b
Veterans Affairs Health Care System, Palo Alto, CA, USA
c
Technische Universität Dresden, Germany
abstractarticle info
Article history:
Received 2 May 2008
Received in revised form 17 July 2008
Accepted 18 July 2008
Available online 25 July 2008
Keywords:
Anxiety
Hyperventilation
Hypoventilation
End-tidal PCO
2
Transcutaneous PCO
2
Recent studies have shown that end-tidal PCO
2
is lower during anxiety and stress, and that changing PCO
2
by
altering breathing is therapeutic in panic disorder. However, end-tidal estimation of arterial PCO
2
has
drawbacks that might be avoided by the transcutaneous measurement method. Here we compare
transcutaneous and end-tidal PCO
2
under different breathing conditions in order to evaluate these methods
in terms of their comparability and usability. Healthy volunteers performed two hypoventilation (slow
vs. paused breathing) and two hyperventilation tests (25 mm Hg at 18 vs. 30 breaths per minute). Three
measurements of PCO
2
(two end-tidal and one transcutaneous device), tidal volume, and respiration rate were
recorded. Before and after each test, subjects filled out a symptom questionnaire. The results show that PCO
2
estimated by the two methods was comparable except that for transcutaneous measurement registration of
changes in PCO
2
was delayed and absolute levels were much higher. Both methods documented that paused
breathing was effective for raising PCO
2
, a presumed antidote for anxious hyperventilation. We conclude that
since the two methods give comparable results choosing between them for specific applications is principally a
matter of whether the time lag of the transcutaneous method is acceptable.
© 2008 Elsevier B.V. All rights reserved.
1. Introduction
Interest in hyperventilation as a cause of anxiety and somatic
complaints has stimulated the measurement of PCO
2
(partial pressure
of CO
2
) inside and outside the laboratory. If breathing is too fast or
deep, more PCO
2
is eliminated than is being produced, resulting in
lower than normal PCO
2
, i.e., hyperventilation (HV). Chemical changes
associated with HV produce physiological symptoms and sometimes
anxiety, either directly or because of the person's fear of bodily sen-
sations (Ley, 1985). Anxiety can lead to more HV, more symptoms, and
more anxiety in a vicious circle. Recent studies have confirmed that
end-tidal PCO
2
is lower during anxiety and stress, and that changing
PCO
2
by altering breathing is therapeutic in panic disorder (PD).
Hypocapnia (low PCO
2
) is often present in PD patients tested while
sitting quietly (e.g., Papp et al., 1997; Wilhelm et al., 2001), and Meuret
et al. (Meuret et al., 2008) found that PCO
2
feedback-assisted
respiratory training was remarkably effective in treating PD. Unlike
in many trials of breathing training for therapeutic purposes, Meuret et
al. measured the actual anti-hyperventilatory effects of the training,
which gave patients feedback of end-tidal PCO
2
levels. Without PCO
2
measurements, the power of breathing instructions to change PCO
2
can easily be overestimated, as demonstrated by the study of Conrad
et al. (Conrad et al., 2007) where instructions commonly recom-
mended for raising pCO
2
failed to do so. Portable, ambulatory devices
have been used in the natural environment to test for a relationship
between hyperventilation and spontaneous panic attacks (e.g., Hibbert
and Pilsbury, 1988; Garssen et al., 1996). The availability of devices able
to monitor PCO
2
outside of the laboratory opened the way for Meuret
et al.'s patients to take home feedback devices for daily practice.
Since PCO
2
is a reflection and a potential cause of anxiety, it is an
important variable in psychophysiological experiments. The study
reported here was motivated by observations of difficulty that many of
our experimental subjects, especially those diagnosed with PD, had in
raising their end-tidal PCO
2
. Normal subjects often fail to raise their
end-tidal PCO
2
significantly in spite of instructions to breathe slowly
and shallowly (Conrad et al., 2007), and certain PD patients fail to do
so in spite of hours of documented practice with a device that feeds
back end-tidal PCO
2
. We wondered whether these failures to raise
PCO
2
could be an artifact of the end-tidal measurement method.
The investigator or clinician who wants to measure this essential
aspect of respiration has two alternatives besides the definitive, direct
measurement of the CO
2
in arterial blood, which unfortunately re-
quires the painful and somewhat risky procedure of placing a sterile
cannula in an artery. One alternative is measuring end-tidal PCO
2
in the
expired air, either by using a mouthpiece, mask, or tent that traps all
International Journal of Psychophysiology 71 (2009) 103–108
⁎ Corresponding author. VA Palo Alto Health Care System (116F-PAD), 3801 Miranda
Ave., Palo Alto, CA 94304, USA. Tel.: +1 650 493 5000x65497; fax: +1 650 493 4901.
E-mail address: eileenwollburg@gmail.com (E. Wollburg).
0167-8760/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijpsycho.2008.07.011
Contents lists available at ScienceDirect
International Journal of Psychophysiology
journal homepage: www.elsevier.com/locate/ijpsycho