Re: Effect of longer term modest salt reduction on blood pressure: Cochrane systematic review and meta-analysis of randomised trials
In their recently published meta-analysis (1), He et al. compares their analysis with ours (2). In contrast to the claim by He et al, the main difference between the two analyses is not the result of the outcomes, but the interpretation of the outcomes.
In 1998 we published a meta-analysis including 83 references on the effect of sodium reduction, which analyzed other outcome variables than blood pressure (2). Requested by the Cochrane institution we have updated our analysis in 2003 (96 references) (3) and 2011 (167 references) (4).
Each of our analyses have been followed by a similar meta-analysis by He and MacGregor, being copies with respect to outcomes and extracted data, but with more restrictive inclusion criteria on study duration and dose of sodium reduction. The first was published in 2002 (25 references) (5). The second was published also as a Cochrane review (28 references) (6). It is, however, surprising that the editors of BMJ have chosen to give a high priority (15 pages) to a third copy by He et al. with a minimal update of 2 references and no news value (1).
He et al. only include studies lasting 4 weeks or longer and with no additional treatment in any of the randomized groups. They argue that this makes a difference. It does concerning the power of the analysis (30 references vs. 167 references), but not convincingly concerning the outcomes. They find the effect on the blood pressure in hypertensive whites to be −5.12 mm Hg/ −2.66 mm Hg, which is similar to our finding in studies lasting more than 4 weeks (-5.18/-2.59). Although the effect on the blood pressure in normotensive whites was a little big bigger than our estimate, the difference between the two analyses was still small (-2.11/-0.88 vs. -1.29/-0.45). Like in our analysis, He et al. also find statistically significant increases in renin, aldosterone and noradrenalin, but they do not find effects on the lipids, which we also did not in studies lasting 4 weeks or more. However only 5-8 studies (262-365 participants) are included in the 4 lipid analyses, which have p-values between 0.11 and 0.22, indicating that the lack of significance could be due to lack of power. Consequently it is too early to decide whether the significant effect on lipids we found in 19-25 studies (586-773 participants) with duration of 1 week or more is temporary or permanent.
In their conclusion He et al. link the association between blood pressure and mortality and between sodium reduction and blood pressure to conclude that sodium reduction is associated with mortality. This is, however, a hypothesis not proven by scientific data. Such a link is not obligatory, as exemplified by the effect of beta blockers, which reduce blood pressure in hypertensives twice as much as sodium reduction, but do not reduce mortality in a meta-analysis of 23613 participants (7). The reason for this is unknown, but could be side effects. However, in spite of the significant side effects on hormones and borderline significant trends on lipids in their own analysis, He et al. explains away this possibility and they also denies the possibility of a J-shaped association between salt intake and mortality shown in several recent population studies. Instead they rely completely on the unproven blood pressure hypothesis. With these firm beliefs He et al. can - in their copy paper – repeat unjustified conclusions from innumerable previous papers.
In contrast, we think that general recommendations based on one single outcome (a small blood pressure effect) without knowing the consequences of this effect and possible accompanying side effects on morbidity and mortality is potentially hazardous.
1) He FJ, Li L, MacGregor GA. Effect of longer term modest salt reduction on blood pressure: Cochrane systematic review and meta-analysis of randomised trials. BMJ 2013;346:f1325 doi: 10.1136/bmj.f1325
2) Graudal NA, Galløe AM, Garred P. Effects of sodium restriction on blood pressure, renin, aldosterone, catecholamines, cholesterols, and triglyceride: a metaanalysis. JAMA 1998; 279:1383–1391.
3) Jürgens G, Graudal NA.Effects of low sodium diet versus high sodium diet on blood pressure, renin, aldosterone, catecholamines, cholesterols, and triglyceride. Cochrane Database Syst Rev. 2003;(1):CD004022.
4) Graudal NA, Hubeck-Graudal T, Jurgens G. Effects of low sodium diet versus high sodium diet on blood pressure, renin,aldosterone, catecholamines, cholesterol, and triglyceride. Cochrane Database Syst Rev. 2011 Nov 9; (11):CD004022. doi:10.1002/14651858.CD004022.pub3.
5) He FJ, MacGregor GA. Effect of modest salt reduction on blood pressure: a meta-analysis of randomized trials. Implications for public health. J Hum Hypertens. 2002;16:761-70.
6) He FJ, MacGregor GA. Effect of longer-term modest salt reduction on blood pressure. Cochrane Database of Systematic Reviews 2004, Issue 1. Art. No.: CD004937. DOI: 10.1002/14651858.CD004937.
7) Wiysonge CS, Bradley HA, Volmink J, Mayosi BM, Mbewu A, Opie LH. Beta-blockers for hypertension. Cochrane Database Syst Rev. 2012 Nov 14;11:CD002003. doi:10.1002/14651858.CD002003.pub4.
Competing interests: No competing interests