Behavioral Treatment Planning 259biofeedback may not work through a common mechanism, at
least for a subset of patients.
Comorbid Psychological Distress or Disorder
The psychological status of the patient deserves special
attention in order to identify conditions (mood and anxiety
disorders, formal thought disorder, certain personality disor-
ders) that might interfere with treatment and that need to be
handled prior to or concurrent with treatment of the headache
(see Holroyd, Lipchik, & Penzien, 1998; Lake, 2001;
Merikangas & Stevens, 1997; Radat et al., 1999; see also the
chapter by O•Callahan, Andrews, & Krantz in this volume;
and the chapter by Jason & Taylor in this volume). These
authors speculate that attention to comorbid conditions may
be crucial to the success of bothpharmacologic and nonphar-
macologic therapies for certain patients. This conclusion is
based on studies revealing the following:
1.The risk for major depression and anxiety disorders is
higher for migraineurs than for nonmigraineous controls.
2.This in”uence is bi-directional. Migraine increases the risk
of a subsequent episode of major depression (adjusted rel-
ative risk4.8), and major depression increases the risk
of subsequent migraine (adjusted relative risk3.3).
3.Comorbid anxiety and depression lead to increases in dis-
ability and contribute to headaches becoming intractable.
4.Psychological distress is greater in headaches that are
more frequent and chronic.
5.Depression is implicated in the transformation of episodic
to chronic tension-type headache.
6.Certain personality disorders reveal a higher incidence of
headache than otherwise would be expected.
Further evidence for the importance of considering psy-
chological factors is obtained from research that has
attempted to identify variables associated with outcome. For
example, studies have consistently shown that patients dis-
playing only minor elevations on a scale commonly used to
assess depression (Beck Depression Inventory) have a dimin-
ished response to self-regulatory treatments (Blanchard et al.,
1985; Jacob, Turner, Szekely, & Eidelman, 1983) and even
abortive medication (Holroyd et al., 1988). Other variables
(anxiety, scales 1, 2, and 3 of the MMPI) have been suggested
as predictive of response to behavioral treatments as well
(Blanchard et al., 1985; Werder, Sargent, & Coyne, 1981).
Holroyd et al. (1988) found that patients who were high in
trait anger, and to a lesser extent, depressive symptoms, were
less likely to respond to abortive pharmacological agents for
migraine headache but these variables were uncorrelated
with response to a combination of relaxation training and
thermal biofeedback, suggesting that the presence of the trait
anger or depression could indicate nonpharmacological inter-
ventions as a “rst line treatment. Jacob et al. (1983) found
that headache patients without signi“cant depressive sypto-
matology responded better to relaxation training than those
with depressive symptomatology. These data suggest that a
combination of pharmacological and nonpharmacological in-
terventions may be useful, such as nonpharmacological man-
agement of headache combined with pharmacological
management of depression. CBT, which has received exten-
sive support for treating anxiety and depression, may be more
useful when comorbid conditions are present. Finally, signif-
icant reductions in anxiety and depression typically occur
following behavioral treatment, regardless of the headache
type or the extent of headache relief (Blanchard et al., 1986;
Blanchard, Steffek, Jaccard, & Nicholson, 1991).Environmental FactorsIt is also important to be mindful of environmental factors/
consequences that may be serving to maintain pain, as pointed
out long ago by Fordyce (1976). Fowler (1975) has applied
this perspective to headache patients. A patient is most likely
to •learnŽ pain behavior when (a) pain behavior is posi-
tively reinforced or rewarded, or (b) •wellŽ behavior is insuf-
“ciently reinforced, punished, or aversive. Therapists can
unwittingly become a part of the learned pain behavior process
in several different ways. Attention from others is a near
universal reinforcer; the sympathetic ear of a therapist can be
especially powerful. Medication prescribing practices can fos-
ter untoward learning effects as well. Palliative medications
are often prescribed on an •as-neededŽ basis, accompanied by
the caution, •Take this only when you really need it; it is pow-
erful and may be addicting.ŽWhen instructed in this manner,
many patients will delay taking the medication until their pain
becomes barely tolerable or near maximum level. If the med-
ication effectively relieves the headache, medication-taking
behavior has become strongly reinforced and is likely to
become more frequent in the future (based on principles of
learning theory). Similar factors come into play when treating
patients whose headache severity has markedly compromised
their day-to-day functioning (a common occurrence with post-
traumatic headache). Such patients are typically instructed,
•Do only what you canŽ or continue activities •until the pain
becomes unbearable.Ž The patient begins an activity, experi-
ences increased pain, and then stops. Stopping the activity
reduces discomfort and makes the patient less likely to engage
in activity in the future. Consequently, therapists need
to probe for environmental conditions, including familial