Background. The aim of this study was to determine the predictors of p
ersistent palpitations and continued medical utilization in a sample o
f medical patients referred for ambulatory electrocardiographic monito
ring. Methods. A prospective telephone follow-up was conducted with pa
tients who had undergone ambulatory electrocardiographic monitoring 3
months earlier. At inception, patients completed in-person interviews
and self-report questionnaires, assessing somatization, hypochondriaca
l attitudes, bodily amplification (high degree of sensitivity to bodil
y sensations), and two types of life stress (minor daily irritants and
major life changes). At Follow-up, patients completed a structured in
terview about their clinical course, palpitations, and utilization of
medical care during the interval. Results. At 3-month follow-up, 55 of
the inception cohort of 67 patients were interviewed again. The mean
severity of palpitations for the entire sample declined significantly,
but 46 (83.6%) patients continued to experience their presenting symp
toms. Stepwise multiple linear regression revealed that the interactio
n of bodily amplification and daily life stress at inception uniquely
explained 10.0% of the variance in palpitation severity at follow-up.
A four-step model composed of these two interaction terms and age and
education level accounted for 21.4% of the variance in palpitations. T
he medical utilization findings are complementary in that the interact
ion of amplification and daily irritants at baseline predicted the num
ber of unscheduled medical visits over the subsequent 3 months. The to
tal number of ventricular premature contractions occurring during ambu
latory monitoring was nor a significant predictor of palpitations. Con
clusions. Palpitations are more persistent in persons who are both hig
hly sensitive to bodily sensations and who experience a greater number
of minor daily irritants. The existence of either predictor alone is
not sufficient to perpetrate this functional somatic symptom; it requi
res the combination of these predictors.