The Cochrane Review of Sodium and Health
“There is still insufficient power to exclude clinically important effects of reduced dietary salt on mortality or cardiovascular disease (CVD) morbidity”.1 This clear statement summarizes the Cochrane Collaboration review and analysis of seven studies, with 6,257 participants and 665 deaths (293 CVD). This scholarly review not only provides a timely update of a previous report, but also challenges the assumptions of a relation between sodium intake and CVD, and its policy implications.
Cochrane Collaboration reviews are justly respected for their integrity, comprehensiveness, and rigorous application of precise methodology. But, like every meta-analyses, this one depends upon the raw material surveyed. Each of the seven studies in this analysis differed in populations sampled, study design, methods, and scope of available data. In fact, only one of the seven studies sought to assess the health impact of different sodium diets; it was a well-designed, randomized clinical trial with sufficient power to test its primary objective.2 The goal was to determine, in heart failure patients, whether a daily dietary sodium intake of 1.8 or 2.7g (80 vs. 120mEql/Na+/daily, 23mg = 1mEql of Na+) produced superior cardiovascular protection. Medications, including high dose diuretic use, and fluid intakes were standardized to ensure that only sodium intake differed between the two groups. Random allocation to either placebo or supplemental sodium tablets achieved the higher sodium in the control group; careful monitoring confirmed protocol adherence. Death and hospitalization, the predefined end points, were reduced by 25% in those assigned the 2.7-g sodium group. This report provides solid evidence that sodium restriction can cause harm—at least in this one particular setting. The findings have been confirmed by two subsequent studies by the same group.3,4 Although this study indicates the inverse association of higher mortality and hospitalization with sodium intake, its …
Michael H. Alderman (michael.alderman{at}einstein.yu.edu)






