cretion in each trial during follow-up and the corresponding net reduction in systolic (p = 0.004) and diastolic (p = 0.003) blood pressure. At higher levels of baseline 24-hour urinary sodium and of change in 24-hour urinary sodium, change in 24-hour urinary potassium showed a dose-response relationship with effect size for both systolic and diastolic blood pressure (p < 0.01). A similar graded response between change in 24-hour urinary potassium and effect size was observed at higher levels of 24-hour urinary sodium as follow-up for systolic (p < 0.01) but not for diastolic (p = 0.2) blood pressure (Whelton et al., 1997). This finding in the meta-analysis was evident in two 2 × 2 factorial trials (Chalmers et al., 1986; Skrabel et al., 1981). In both trials, supplemented potassium lowered blood pressure when sodium intake was high, but not when sodium intake was low.
The role of urinary sodium excretion as an effect modifier for the relationship between potassium consumption and blood pressure is consistent with results from observational investigations where blood pressure is more closely related to the ratio of urinary sodium:potassium excretion than to either urinary sodium or potassium excretion alone (Khaw and Barrett-Connor, 1988, 1990).
Treatment-related systolic blood pressure effect size estimates were significantly (p = 0.03) greater for the six trials with greater than 80 percent African-American participants compared with the 25 trials with greater than 80 percent white participants (Whelton et al., 1997). Also, there was some evidence for a dose-response relationship between potassium dose and blood pressure and some evidence for greater blood pressure reduction in African-American compared with white participants. In the two trials included that enrolled exclusively African-American individuals, potassium significantly lowered both systolic and diastolic blood pressure (Brancati et al., 1996; Obel, 1989). The blood pressure reductions in the study by Obel (1989) were particularly striking.
Overall, available evidence from observational studies, clinical trials, and meta-analyses of trials documents that higher intakes of potassium lower blood pressure. Blood pressure reductions from supplemental potassium occurred when baseline intake was low (e.g., 32 to 35 mmol/day in Brancati et al., 1996) and when baseline intake was much higher (> 80 mmol/day in Naismith and Braschi, 2003). Because virtually all trials used potassium chloride supplements, while observational studies assessed dietary potassium intake from foods (paired with nonchloride anions), the effects of potassium on blood pressure appear to result from potassium rather than its conjugate anion.