Effect of modest salt reduction on blood pressure: a meta-analysis of randomized trials. Implications for public health.

Two recent meta-analyses of randomised salt reduction trials have concluded that there is little purpose in reducing salt intake in the general population. However, the authors, as with other previous meta-analyses, included trials of very short duration (eg 1 week or less) and trials of acute salt loading followed by abrupt reductions to very Iow salt intake (eg from 20 to less than 1g of salt/day). These acute salt loading and salt depletion experiments are known to increase sympathetic tone, and with salt depletion cause a rise in renin release and, thereby, plasma angiotensin II. These trials are not appropriate, therefore, for helping to inform public health policy, which is for a more modest reduction in salt intake, ie, from a usual intake of &ap;10 to &ap;5g of salt per day over a more prolonged period of time. We carried out a meta-analysis to assess the effect of a modest salt reduction on blood pressure. Our data sources were MEDLINE, EMBASE, Cochrane library, CINAHL, and the reference lists of original and review articles. We included randomised trials with a modest reduction in salt intake and a duration of 4 or more weeks. Meta-analysis, meta-regression, and funnel plots were performed. A total of 17 trials in hypertensives (n=734) and 11 trials in normotensives (n=2220) were included in our study. The median reduction in 24-h urinary sodium excretion was 78 mmol (equivalent to 4.6g of salt/day) in hypertensives and 74 mmol in normotensives. The pooled estimates of blood pressure fall were 4.96/2.73 ± 0.40/0.24 mmHg in hypertensives (P<0.001 for both systolic and diastolic) and 2.03/ 0.97 ± 0.27/0.21 mmHg in normotensives (P<0.001 for both systolic and diastolic). Weighted linear regression analyses showed a dose response between the change in urinary sodium and blood pressure. A reduction of 100 mmol/day (6g of salt) in salt intake predicted a fall in blood pressure of 7.11/3.88 mmHg (P<0.001 for both systolic and diastolic) in... [ABSTRACT FROM AUTHOR]