Is the association of hypertension and panic disorder explained by clustering of autonomic panic symptoms in hypertensive patients?

Background
Autonomic nervous system dysfunction may be implicated in the association of hypertension with panic attacks and panic disorder. We hypothesised that panic symptoms of autonomic origin are more common in attacks experienced by hypertensive than normotensive patients, that autonomic panic symptoms cluster together as a distinct factor, and that this factor is more prevalent in hypertensive patients with panic than in normotensives.

Methods
We analysed all 346 structured questionnaires completed by primary care and hospital clinic patients who had reported experiencing full (n = 287) or limited symptom panic attacks (n = 59) (268 with hypertension, and 78 never having had hypertension). Frequency of sweating, flushes, and racing heart, symptoms selected prospectively as being most likely of autonomic origin, were compared between hypertensive and normotensive patients. Principal component analysis was performed with varimax orthogonal rotation. Using logistic regression, odds ratios were calculated for association of factor scores with hypertension.

Results
Sweating and flushes were significantly more common among hypertensive patients than normotensives (sweating; 65% v 46%, p = 0.003, flushes; 55% v 40%, p = 0.019). There was no significant difference between groups for frequency of racing heart nor any of the remaining panic symptoms analysed as secondary endpoints. Principal component analysis yielded four factors with eigenvalues > 1.0. Factor 1 was dominated by autonomic symptoms, notably sweating and flushes, which had loadings of 0.68 and 0.61. On regression only this autonomic factor showed a significant association with hypertension, the odds ratio being 1.37 (95% C.I. 1.05 to 1.77, p = 0.018).

Conclusions
These findings support the possibility that autonomic dysfunction contributes to the association of hypertension with panic.