The use of Fourier analysis in the calculation of trough to peak ratio from ambulatory blood pressure measurements

Abstract

The trough:peak (T:P) ratio serves as an index of efficacy of antihypertensive drugs with respect to their dose and dose interval. There is no consensus regarding the method for the calculation of the T:P ratio. We assessed the influences of curve smoothing, the average fall in 24-h mean arterial pressure (MAP) and the length of the peak effect period on the result of T:P ratio calculation. Forty-two patients with essential hypertension (aged 27–81 years; 20 males) had a 24-h ambulatory blood pressure (BP) measurement on two occasions. The first was performed at baseline, the second after 12 weeks of treatment with a β-blocker, angiotensin-converting enzyme inhibitor, calcium slow-channel blocker, diuretic or a centrally acting drug, all taken once daily. BP data were analysed both by Fourier analysis (FA) with four harmonics and by time block analysis (TBA). The peak effect was defined as the maximum drop in MAP over a period of 0 to 3 h following drug intake at any time in the 24 h, and the trough effect as the fall in MAP over the last 2 h of the dose interval. FA gave higher T:P ratio values than TBA (0.51 vs 0.43; P < 0.05) and the variability of the results was reduced by fa (fa: 0.49–0.52; tba: 0.35–0.46). a greater fall in 24-h map was associated with a higher t:p ratio (r = 0.42, P < 0.02; tba: r = 0.31, P = 0.09). Half of the patients whose fall in 24-h MAP after treatment was <10 mm hg, had higher trough bp values, resulting in negative t:p ratios. with increasing length of the peak period, the calculated peak effect was reduced and the t:p ratio increased (both fa and tba: P < 0.001). the mean trough responses during the last 2 h of the dose interval were similar after tba and fa.based on these data, we recommend the use of fa for calculation of t:p ratios, especially when the magnitude of the antihypertensive effect is small. we also suggest using peak- and trough-effect periods of 2 h each. whenever t:p ratios are calculated, the method used for curve smoothing and the length of peak- and trough-effect periods should be specified in order to obtain meaningful results. t:p ratios need to be interpreted individually together with the magnitude of the antihypertensive drug effect and the circadian bp profile.