Behavioral Treatment 253In addition to meta-analytic approaches, various groups
have assembled panels to conduct evidence-based reviews,
wherein rigorous methodological criteria are used to evaluate
every study under consideration. Evidence-based analyzes
have been performed by the Division 12 Task Force, the U.S.
Headache Consortium (composed of the American Academy
of Family Physicians, American Academy of Neurology,
American Headache Society, American College of Emer-
gency Physicians, American College of Physicians-American
Society of Internal Medicine, American Osteopathic Associ-
ation, and National Headache Foundation) (Campbell,
Penzien, & Wall, 2000), and the Task Force of the Society of
Pediatric Psychology (Holden, Deichmann, & Levy, 1999).
Consideration of the “ndings from these studies leads to
the following conclusions. First, relaxation, biofeedback, and
cognitive therapy lead to signi“cant reductions in headache
activity, ranging from 30% to 60%. Second, conversely, there
are a fair number of patients who are nonresponders or partial
responders (approximately 40% to 70%). Prediction of
treatment response and careful treatment planning become
particularly important when attempting to improve on this
outcome. Upon their extensive review, the U.S. Headache
Consortium concluded that behavioral treatments may be
particularly well suited for patients having one or more of
the following characteristics: The patient prefers such an ap-
proach; Pharmacological treatment cannot be tolerated or is
medically contraindicated; The response to pharmacological
treatment is absent or minimal; The patient is pregnant, has
plans to become pregnant, or is nursing; The patient has a
long-standing history of frequent or excessive use of anal-
gesic or acute medications that can aggravate headache; or
The patient is faced with signi“cant stressors or has de“cient
stress-coping skills. More is said about treatment prediction
later. Third, improvements exceed those obtained for various
control conditions. Fourth, nonpharmacological treatments
produce bene“ts similar to those obtained for pharmacologi-
cal treatments. Fifth, combining treatments often increments
effectiveness, especially so for nonpharmacological and
pharmacological. However, the net gain of adding a second
treatment modality beyond a single treatment sometimes is
relatively small. This again stresses the importance of “nding
the •rightŽ therapy or combination of therapies for an indi-
vidual patient. Research into the prediction of treatment re-
sponse may elucidate some of this and allow clinicians to
maximize therapeutic gains. Sixth, most studies of nonphar-
macological interventions have included subjects that contin-
ued their consumption of any number of pharmacological
agents while undergoing nonpharmacological interventions.
Only a very few studies have systematically isolated pure
treatments (e.g., Holroyd et al., 1988, 1995, 2001; Mathew,
1981; Reich, 1989).
There is also a fair amount of evidence to suggest that the
effects of these types of therapies are durable. A number of
studies have found substantial maintenance of treatment
gains, at least among those who respond initially, for periods
of up to seven years posttreatment (see Blanchard, 1992), and
that these effects maintain whether further contact is pro-
vided (booster sessions) or not (Andrasik, Blanchard, Neff, &
Rodichok, 1984). For example, in a prospective follow-up,
Blanchard, Appelbaum, Guarnieri, Morrill, and Dentinger
(1987) found that 78% of tension headache sufferers and
91% of migraine headache sufferers remained signi“cantly
improved (as assessed by headache diary) “ve years follow-
ing completion of relaxation training and/or biofeedback
training. In a retrospective four-year follow-up study of
almost 400 headache patients who had completed a compre-
hensive clinical program including several types of biofeed-
back and relaxation training, Diamond and Montrose (1984)
found that 65% reported maintenance of treatment gains.
While this latter study is retrospective, the results are encour-
aging because of the very large sample size and the fact that
the data were collected from patients enrolled in a clinical
program (as opposed to a research program). As such, these
data may provide information about follow-up with •natural-
isticŽ or •real lifeŽ clinical programs.BEHAVIORAL TREATMENTA Biobehavioral Model of HeadacheThe biobehavioral model, which guides treatment of head-
ache, states that the likelihood of any individual experiencing
headache depends on the speci“c pathophysiological mecha-
nisms that are •triggeredŽ by the interplay of the individual•s
physiological status (e.g., level of autonomic arousal), envi-
ronmental factors (e.g., stressful circumstances, certain
foods, alcohol, toxins, hormonal ”uctuations), the individ-
ual•s ability to cope with these factors (both cognitively and
behaviorally), and consequential factors that may serve to
reinforce, and thus increase, the person•s chances of report-
ing head pain (Martin, 1993; Waggoner & Andrasik, 1990).
The main determinant for the resulting headache is the patho-
physiological biological response system that is activated.
Psychological and behavioral factors do not play a causal role
per se. Rather, they contribute to headache as factors that
(a) trigger, (b) maintain, or (c) exacerbate headache, or (d) as
sequelae to continued head pain that subsequently disrupt
overall functioning.