New salt-sensitivity metrics: variability-adjusted blood pressure change and the urinary sodium-to-creatinine ratio.

Design: We report the results of a 24-week, placebo-controlled, two-period, crossover trial of sodium supplementation in 112 normotensive African Americans, aged 25 to 64 years, with average blood pressure (BP) of 105/70 mm Hg. Estimated 24-hour urinary sodium excretion was 133.6 mmol; the average urinary sodium-to-creatinine ratio was 0.74.
Methods: Variability-adjusted BP change was the difference in BP level after the respective treatment periods, divided by the intra-person standard deviation of the average BP obtained at 3 consecutive screening visits during a 4-week period.
Results: The urinary sodium-to-creatinine ratio and the total urine sodium content were 37.8% and 26.5% higher, respectively, at the end of the sodium treatment period. Twenty-four hour ambulatory BP change (mm Hg) (95% CI) was systolic 1.2 (0, 2.4), and diastolic 0.7 (-0.3, 1.8); cuff BP change was systolic 0.9 (-1, 2.9), and diastolic 1.4 (0.1, 2.7). Variability-adjusted BP change was systolic 0.2 (-0.4, 0.8) and diastolic 0.4 (-0.1, 0.9). Though variability-adjusted and unadjusted SBP change correlated highly (r = 0.941, P<.001), only the former correlated with body mass index (r = 0.224, P<.05), a known correlate of salt sensitivity. While total urinary sodium content in timed urine collections and urinary sodium-to-creatinine ratio correlated (r = 0.727, P<.001), neither correlated with cuff BP changes. Change in urinary sodium-to-creatinine ratios of 3 consecutive pooled overnight 8-hour urine collections correlated with changes in 24-hour ambulatory SBP (r = 0.294, P<.001) and DBP (r = 0.193, P<.05); however, change in total urinary sodium content was uncorrelated. Total urinary sodium content of these pooled collections (P = .001), but not the urinary sodium-to-creatinine ratio, was positively related to urinary creatinine excretion per kilogram of body weight, the latter being an indicator of urine collection duration.
Conclusions: The lack of effect of the duration of urine collection on the urinary sodium-to-creatinine ratio is advantageous in individuals who may report inaccurately the duration of their urine collection. Sequential regression analyses demonstrated that the urinary sodium-to-creatinine ratio conveyed all of the changes in urinary sodium excretion information contained in aggregate urinary sodium excretion-and more. Variability-adjusted BP change was the more sensitive metric of BP response to dietary sodium manipulations, than unadjusted BP change. Thus, variability-adjusted BP change and the urinary sodium-to-creatinine ratio appear to be incrementally better metrics of salt sensitivity than those traditionally used.