Progress in Cardiovascular Diseases 61 (2018) 20–26

Contents lists available at ScienceDirect

Progress in Cardiovascular Diseases

journal homepage: www.onlinepcd.com

Conflicting Evidence on Health Effects Associated with Salt Reduction

Calls for a Redesign of the Salt Dietary Guidelines☆
Niels Graudal a,⁎, Gesche Jürgens b
a
Department VRR 4242, Copenhagen University Hospital, Rigshospitalet, Denmark
b
Clinical Pharmacology Unit, Zealand University Hospital, Roskilde, Denmark

a r t i c l e i n f o a b s t r a c t

Article history: Ninety-five percent of the World's populations have a mean salt intake between 6 and 12 g, which is much lower
28 April 2018 than the tolerated daily level of up to 55 g/d. In spite of this, the recommended upper level by many health insti-
28 April 2018 tutions is as low as 5.8 g/day. When reviewing the evidence for an upper level of 5.8 g/day, it becomes apparent
that neither the supporting studies selected by the health institutions, nor randomized controlled trials and pro-
Keywords:
spective observational studies disregarded by the health institutions, document that a salt intake below this 5.8 g,
Salt
Blood pressure
has beneficial health effects. Although there is an association between salt intake and blood pressure, both in ran-
Renin-angiotensin-aldosterone domized controlled trials and in observational studies, this association is weak, especially in non-obese individ-
Lipids uals with normal blood pressure. Furthermore a salt intake below 5.8 g is associated with the activation of the
And mortality renin-angiotensin-aldosteron system, an increase in plasma lipids and increased mortality. A redesign of the
salt dietary guidelines, therefore, seems to be needed.
© 2018 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).

Contents

Abbreviations: AHA, American Heart Association; BMI, body mass index; BP, blood pressure; CDC, Centers for Disease Control and Prevention; CVD, cardiovascular disease; DASH, the
Dietary Approaches to Stop Hypertension; FDA, Food and Drug Association; g/d, gram per day (24 h); HF, heart failure; HTN, hypertension; IOM, Institute of Medicine; Na, sodium
(natrium); NAM, National Academy of Medicine; NHLBI, National Heart, Lung and Blood Institute; NIH, National Institute of Health; RAAS, renin-aldosterone-angiotensin system; RCTs,
randomized controlled trials; SR, salt reduction; WASH, World Action on Salt and Health; WHO, World Health Organization.
☆ Statement of conflict of interest: see page 25.
⁎ Corresponding author at: Department VRR 4242, Copenhagen University Hospital, Rigshospitalet, Blegdamsvej 9, DK-2100 Copenhagen, Denmark.
E-mail address: graudal@dadlnet.dk (N. Graudal).

https://doi.org/10.1016/j.pcad.2018.04.008
0033-0620/© 2018 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
21 N. Graudal,
N. Graudal,
G. Jürgens
G. Jürgens
/ Progress
/ Progress
in Cardiovascular
in Cardiovascular
Diseases
Diseases
61 (2018)
61 (2018)
20–26
20–26 21

Introduction
Ninety-five percent of the World's populations have a mean salt in-
take between 6 and 12 g/d,1 the minimum required amount being about
0.5 g/d.2 Salt intake of up to about 55 g/d has been recorded.3 Random-
ized controlled trials (RCTs) of salt intake in the interval 0.5–40 g/d have
not reported deficiency or toxic symptoms.4 Intoxication has been de-
scribed after rapid intake of about 50 g or more over a few minutes.5
Salt is essential for life, as it contributes to the action potentials and
membrane potentials of cells and maintains extracellular volume and
blood pressure (BP).6 Centers in the brain regulate body salt, together
with the renin-aldosterone-angiotensin system (RAAS) and the
kidneys.6,7
Still, many health institutions agree that salt is as toxic as tobacco
and, therefore, consider salt to be a target for prevention.8,9 This position
is based on a belief that salt not only maintains BP, but increases BP as a
linear function of the ingested amount, leading to increased mortality.8
In recent years a significant number of RCTs and population studies
have questioned the harmful effects of salt.4,10 Many health institutions
disagree and still support interventions to reduce salt intake in the gen-
eral population to below 5.8 g8,9,11 in parallel with attempts to reject the
outcomes of studies showing harmful effects of low salt intake.12 Repre-
sentatives from the American Heart Association (AHA) reviewed the
methodological quality of 26 population studies to analyze whether
methodological issues accounted for the lack of beneficial effects of
low salt intake. They concluded that these studies, due to methodolog-
ical issues, were not suited to form the basis for salt guidelines, which
should instead be based on “the robust body of evidence linking Na
with elevated blood pressure and the few existing general population
trials of the effects of Na reduction on cardiovascular disease (CVD)”.12
However, as indicated in a meta-analysis of RCTs4 this evidence may
not be as robust as previously claimed.12 We therefore find it pertinent
to present a collective critical review of the quality of the key health ev-
idence for salt reduction (SR) promoted by the health institutions in the
context of the existing evidence from RCTs and prospective population
studies.
increase in BP with increased salt intake both below and above a salt in-
take of 5.8 g, thus justifying the conclusion that salt intake should be
below 5.8 g. Fig. 1 shows all known dose-response studies published
at the time of the IOM evaluation. The 3 studies13-15 considered by
IOM were all studies of older individuals with hypertension (HTN)
and those with the steepest dose-response relationships. The HTN
study with the lowest dose-response relationship16 and four normoten-
sive studies without dose-response relationships (Fig. 1)16,17 were not
included in the evaluation. This selective inclusion was further empha-
sized by the fact that high BP and age biased the three selected studies.
The first study included older individuals with very high BP.13 Mem-
bers of the WASH group performed the second study.14 This group has
published 10 studies in individuals with HTN showing a mean effect of
SR on systolic BP(SBP) of about 10 mm Hg, which is twice that reported
in all other RCTs of HTN (4.85 mm Hg) (Table 1). This indicates a sys-
tematic bias, which only partly depends on the high baseline BP of the
included populations. The chair of the 2005 IOM committee7 was co-
author of the third study.15 This study, the Dietary Approaches to Stop
Hypertension (DASH) study, is the most highlighted of the studies
used to justify SR.
The Dietary Approaches to Stop Hypertension (DASH) studies15,18
In the first DASH study18 3 diets were compared, a control diet de-
pleted in potassium, calcium and magnesium to the 25% percentile of
the population in order “to ensure a marked contrast to the ideal dietary
patterns”. The ideal diet was rich in potassium, calcium, magnesium,
carbohydrate, protein and fiber and low in fat. The third diet was in be-
tween these two diets. At the time of the design of this study it was well
known that potassium intake was inversely associated with BP.19,20
Thus the BP-reducing effect of the potassium rich fruit supplemented
DASH diet, identified in this study, may in part be ascribed to the
designed potassium depletion in the control group.
The experiences from the first DASH trial18 were carried forward to
the subsequent DASH sodium trial,15 in which the intermediate diet was
eliminated and the participants were randomized to the ideal and

Methods

We included RCTs and population studies as well as studies based on

RCTs and population studies (meta-analyses and modeling studies) ini-
tiated or economically supported by the following health institutions
and organizations, which give high priority to salt reduction: Centers
for Disease Control and Prevention (CDC), National Institute of Health
(NIH), National Heart, Lung and Blood Institute (NHLBI), Institute of
Medicine (IOM, now National Academy of Medicine (NAM)),
American Heart Association (AHA), Food and Drug Association (FDA),
World Health Organization (WHO), and World Action on Salt and
Health (WASH). The study identification was in part based on our
previous systematic searches of RCTs4 and population studies,10 and
reference lists from reports, studies and position papers published by
the mentioned institutions.

Salt: effect on BP

IOM (now NAM) report on dietary salt reference intake

The upper limit for salt intake of 5.8 g recommended by health insti-
tutions originates from an IOM report,8 which states, “most relevant to
determining an upper level are the three trials in which the lowest
level of dietary sodium intake was close to the adequate intake (Johnson
et al., 2001; MacGregor et al., 1989; Sacks et al., 2001).” “In view of the
results from these three trials, the lowest-observed adverse-effect Fig. 1. Individual study systolic BP response to increasing changes in sodium urinary
excretion (as a measure of sodium intake) in otherwise healthy normotensive and
level for dietary sodium is set at 2.3 g/day (5.8 g salt/day).” These 3 hypertensive individuals. Institute of Medicine7 used studies 13, 14 and 15 to estimate
studies,13-15 which randomized individuals to 3–4 different doses of the 5.8 g upper limit for salt intake *Burnier: J Hypertens 2000; 18:1657-64; *Fuchs:
salt (dose-response analyses), all showed a significant proportional BJMBR 1987;20:25-34; *Heer: AJPRP 2000; 27: 278:F585-95.
22 N. Graudal,
N. Graudal,
G. Jürgens
G. Jürgens
/ Progress
/ Progress
in Cardiovascular
in Cardiovascular
Diseases
Diseases
61 (2018)
61 (2018)
20–26
20–26 22

Table 1
The effect of sodium reduction on systolic and diastolic blood pressure (SBP/DBP) in studies performed by researchers from the World Action on Salt and Health group.

Reference Mean age (years) Baseline SBP/DBP (mm Hg) Duration, (days) Sodium reduction (mmol) Effect SBP/DBP (mm Hg)

1) Lancet 1982;i:351 49 156/98 28 76 −10/−5

2) BMJ 1987;294: 531 52 150/97 30 100 −13/−9
3) J Hypertens 1988;6:613 52 157/101 5 97 −9/−5.6
4) Lancet 1989;II:1244 57 163/100 30 141 −16/−9
5) Hypertension 1991;17:798 54 147/91 30 91 −9/−3
6) J Hypertens 1994;12:809 49 144/100 5 296 −11.6/−5
7) J Hum Hypertens1996;10:523 46 151/96 7 293 −15.2/−3.7
8) Lancet 1997;350:850 67 162/90 30 81 −7.2/−3.2
9) Hypertension 2005;46:308 63 156/100 28 78 −8/−3
10) Hypertension 2009;54:482 50 147/91 42 55 −5/−3
Mean effect of study 1–10, SBP −10.21 [−12.75, −7.67]
Mean effect of study 1–10, DBP −4.30 [−5.68, −2.92]
Mean effect of all hypertensive studies except study 1–10, SBPa −4.85 [−5.80, −3.91]
Mean effect of all hypertensive studies except study 1–10, DBPa −2.67 [−3.27, −2.07]
a
Data obtained from the data file of reference 23.

control diets, respectively, and in addition crossed over to three differ-
ent salt intake diets. SR reduced SBP by 6.7 mm Hg in the potassium de-
pleted control diet and by 3 mm Hg in the ideal diet. A meta-analysis of
five studies identified among 176 salt-reduction studies,4 which in addi-
tion to allocation to a reduced salt diet also allocated participants to a
combination of sodium reduction and potassium supplementation,
shows a significant BP difference between a low sodium/low potassium
diet versus a low sodium/high potassium diet (Fig. 2). This difference
corresponds to the differences observed in the DASH sodium trial, indi-
cating that the planned potassium depletion in the control group am-
plifies the BP reduction induced by salt reduction. Furthermore, the
mean age, body mass index (BMI) and baseline BP of the included par-
ticipants were significantly higher than the average American popula-
tion. This general bias is reflected in a supplementary publication of
the DASH trial, which shows that in younger individuals between 21
and 42 years, representing more than 50% of the American population,
the supplied low-sodium/high potassium diet has no effect on SBP com-
pared with the depleted low-sodium/low potassium diet.21 Due to the
design, interpretation and use of the DASH-sodium trial, it is a major
limitation that the data from this government-funded trial are not
publically available.22
WASH and WHO versus Cochrane: meta-analyses of the effect of reduced
dietary salt intake on BP
Cochrane has published two salt-reduction reviews by two different
author groups measuring exactly the same outcomes4,23; the justifica-
tion being that the original review4 investigates acute effects of SR,
whereas the more recent review from 2004 by members of the WASH
group23 investigates longer-term effects. However, in 2004 there was
no scientific justification for this distinction and a later review of longi-
tudinal RCTs showed no differences in the effect of SR on BP between
week 1 and week 6.17 The WHO review24,25 includes almost the same
studies as the WASH review. Table 2 compares the original Cochrane re-
view, the WASH review and the WHO review. In general there were no
differences between the BP effects verifying that the distinction be-
tween acute and long-term studies is not justified. The marginally
higher effect in the analysis of normotensive studies in the WASH re-
view was due to the effect of the DASH study and 3 studies, which in-
cluded both normotensive individuals and those with HTN (Table 3).
After exclusion of these 4 studies the 8 remaining studies showed an
effect, which was almost identical with the Cochrane review and the
WHO review (Table 2).
CDC/FDA evidence for relation between reduced dietary salt intake and BP
Recently, the CDC and FDA released a proposal for voluntary guide-
lines to encourage food companies to steadily reduce sodium in proc-
essed and restaurant foods.11,26 The argument was “strong evidence,
including a recent analysis of more than 100 randomized clinical trials,
that sodium reduction reduces blood pressure in adults.” However,
this analysis27 was based on 65% HTN studies and 35% normotensive
studies. The meta-regression line with a slope of 3.8 mm Hg per
100 mmol sodium (2.3 g sodium) was forced through zero and was pri-
marily based on data adopted from the original Cochrane review.4 The
appropriate function with a constant reveals that the slope is only
2.27 mm Hg/100 mmol (Fig. 3). The authors applied the no-constant lin-
earity from the mixed meta-regression analysis to both the HTN and the

Fig. 2. Meta-analysis of the effect of low sodium/high potassium diet vs. low sodium/low potassium diet on systolic blood pressure and the effect of low sodium ideal diet (high potassium)
vs. low sodium control diet (low potassium) in the DASH study.15
23 N. Graudal,
N. Graudal,
G. Jürgens
G. Jürgens
/ Progress
/ Progress
in Cardiovascular
in Cardiovascular
Diseases
Diseases
61 (2018)
61 (2018)
20–26
20–26 23

Table 2
Comparison of original Cochrane review (2003–2017), WASH group Cochrane review (2004–2013) and WHO review (2013): blood pressure.

Graudal et al. 2003/2011 (original) MacGregor et al. 2004 (WASH) Aburto et al.
Cochrane 2017 (4) Cochrane 2013 (23) WHO 2013 (25)

Hypertension Normal BP Hypertension Normal BP Hypertension Normal BP#

N (n) 86 (6001) 90(8833) 22 (990) 12(2240) 24 (2273) 7 (3067)

Median Age (range), years 51.6 29 50 50 – –
Median SBP/DBP, mm Hg 151/93 119/71 148/93 127/77 – –
Median duration, weeks 4 1 5 4 4 4
Usual sodium, mean, mmol 183 199 162 153 – –
Low sodium, mean, mmol 80 45 87 78 – –
Sodium reduction, mean, mmol 103 154 75 75 – –
Effect SBP/DBP −5.51/−2.88 −1.09−/0.03 −5.39/−2.82 −2.42/−1 −4.06/−2.26 −1.38/−0.58
Effect SBP/DBP (subgroup) (N,n) −1.31/−0.36 (59, 7125)a −1.63/−0.43 (8, 2113)b −1.38/−0.58 (7, 3067)

N: Number of studies; n: number of participants;

# 5 borderline studies were not included in the analysis.
a
Studies with duration of at least 7 days.
b
DASH study and studies of mixed hypertensive/normotensive individuals excluded.
normotensive individual studies and standardized the systolic BP effect
to 2.3 g (100 mmol). In contrast, a separate meta-regression analysis of
the normotensive studies shows that neither the assumption of linearity
nor the cut point of zero is valid for the normotensive studies (Fig. 3).
Thus the CDC and FDA assumed dose-response relationship is not
valid for the 75% of the population with a normal BP.
CDC/NHLBI analysis of association between salt intake and BP in NHANES 2014
In the main article this recent analysis shows an association between
sodium intake and BP, which is stronger than found in previous popula-
tion studies (4.58/2.25 mm Hg/1 g Na).28 However, according to eTable 2
in their supplement, this effect was mainly due to the adipose 50% of the
population with a BMI above 30. In the group of participants with a BMI
less than 30 the systolic BP effect was only 1.8 mm Hg/1 g Na. This is
similar to the effect found in the worldwide Prospective Urban Rural Ep-
idemiology (PURE) study (2.1 mm Hg/1 g Na), in which the study pop-
ulation had a mean BMI of 26.29
Salt: effect on hormones and lipids
WASH and WHO versus Cochrane: meta-analyses of the effect of reduced
dietary salt intake on hormones and lipids
mean level below 5.0 g in accordance with the recommendations,
whereas the two small analyses of studies of at least 4 weeks
duration23,25 reduce salt intake down to, but not below, a mean level
of 5.0 g corresponding to the WHO recommendations, which are to re-
duce salt below 5.0 g/d.24 Thus, the level of salt intake, rather than the
duration of the exposure, determines the occurrence of side effects.
This has been verified in Yanomamo Indians on low-salt intake, who
have persistently elevated levels of renin and aldosterone in the
blood2; and in a cross-sectional study of individuals with normal BP
and hypertension.7 RAAS is activated by salt reduction in healthy and
sick individuals, but McCarron suggested that in patients with heart fail-
ure (HF) and renal disease, whose RAAS is activated by a compromised
renal perfusion, the activated RAAS might dictate a neural-driven in-
crease in salt intake in an attempt to increase renal perfusion and sup-
press plasma renin activity and angiotensin II and its pathologic
impact on the heart and vasculature.30 Under these circumstances a
high salt intake was suggested not to be causative, but more likely a
compensatory response mechanism, serving as a natural RAAS
inhibitor.30 That interpretation is consistent with recent studies of pa-
tients with HF and renal disease showing that a low salt diet was not as-
sociated with reduced morbidity or mortality.31,32
Salt: effect on health outcomes

The effect of salt-reduction on renin, aldosterone, noradrenalin, RCTs relating salt intake to health outcomes
adrenalin, cholesterol and triglyceride has been analyzed in the three
previously mentioned meta-analyses.4,23,25 The results are presented RCTs measuring health effects have been performed in individuals
in Table 4. The original Cochrane review4 reduces salt-intake to a with HTN and pre-HTN overweight individuals, but not in healthy

Table 3
Studies of participants with normal BP included in WASH Cochrane meta-analysis, 201323

Reference Baseline SBP/DBP Sodium intake, mmol SBP effect mm Hg DBP effect mm Hg
mm Hg High/low

Puska, Lancet 1983;I:1-5. 132/82 192/77 −1.5 (3.32) −2.1 (2.03)

Watt, BMJ 1985;291:1525-8 113/65 130/68 (HH) −1.4 (0.74) 1.2 (0.93)
(LL) −0.5 (0.82) 1.4 (0.9)
Mascioli, Hypertension 1991;17(S1):I21-6 131/84 179/109 −3.6 (0.9) −2.3 (0.8)
TOHP I, JAMA 1992;267:1213-20 125/83 144/100 −1.7 (0.59) −0.9 (0.42)
Cobiac, J Hypertens 1992;10:87-92 134/78 148/79 −1.7 (2.14) 0.8 (1.01)
Ruppert, Hypertens 1993;11:743-9 113/72 200/82 1.7 (2.39) 1 (1.64)
Nestel, Hypertens 1993;11:1387-94 129/77 157/106 (F) −6 (4.9) −2 (3.31
(M) −2(3.43) −1 (2.65)
Schorr, J Hypertens 1996;14:131-5* 132/72 166/105 −7.2 (4.9) −2.9 (2.61)
D12h**: 140/84
Cappucio, Lancet 1997;350:850-4* 149/84 167/91 −8.2 (3.07) −3.9 (1.65)
TOHP II, Arch Intern Med 1997;157:657-67 128/86 178/135 −1.2 (0.5) −0.7 (0.4)
DASH, NEJM 2001;344:3-10* 129/84 141/64 −5.3 (0.77) −2.6 (0.5)
Melander, J Hypertens 2007;25:619-27* 136/78 140/51 −4.6 (2.1) −2.8 (1.03)
D12h**: 141/90

*Excluded in sensitivity analysis; **12 h BP during daytime; HH: Parents high BP; LL: Parents low BP; F: Female; M: Male.
24 N. Graudal,
N. Graudal,
G. Jürgens
G. Jürgens
/ Progress
/ Progress
in Cardiovascular
in Cardiovascular
Diseases
Diseases
61 (2018)
61 (2018)
20–26
20–26 24

Fig. 3. Sodium reduction versus mean reduction of systolic blood pressure (MRSBP). Univariable analysis: Each circle shows the MRSBP outcome of one study comparing a reduced sodium
intake versus a usual sodium intake (closed circles: Normotensive studies; stippled circles: Hypertensive studies). The size of the circle corresponds to its inverse variance weight of the
MRSBP. Regression lines are shown for all studies with recommended method including a constant (y = −0.0227×−2.0374), the no-constant method (y = −0.0382×) and for
normotensive studies only (y = −0.0005×−1.988).
individuals. Collectively they showed a non-significant 24% reduction in although side-effect data were available in the meta-analyses from
CVD events in the low-salt group (data incompletely recorded) and no which the BP data were adopted. These models predicted thousands
difference in all-cause mortality (data completely recorded).33 The to millions of saved lives by dietary salt reduction in contrast to real
mean salt intake in the low-salt groups was 5.8 g or higher. Thus there data from cohort studies, which indicate that low salt intake is associat-
are no RCTs to show health effects of salt-intake below 5.8 g. ed with increased mortality.

Modeling studies relating salt intake to health outcomes Cohort studies relating salt intake to health outcomes

Modeling studies establish a dose-response relationship between
salt intake and BP, which indirectly is used to translate salt reduction
to reduction in mortality by means of data from observational studies
linking BP to mortality. One modeling study34 used the linear regression
analysis based on data from the WASH meta-analysis (Table 3), data
from the DASH study15 (Fig. 1) and data from one of the WASH group
studies14 (Fig. 1) (study no. 4 in Table 1) to construct the dose-
response analysis. Another study27 used the data from the above de-
scribed meta-regression analysis of 103 RCTs of which 65% were HTN.
None of the modeling studies included side-effect data in the models,
Evidence from WHO is based on a meta-analysis from 2009,35 which
was updated in 2013.25 In the 2009 analysis the relative risk of higher
versus lower salt intake was investigated by comparing the event rate
in the two categories with a difference in average salt intake closest to
5.8 g/day. In the updated analysis25 the overall effect estimate was gen-
erated comparing the risk of each outcome in the lowest salt intake
group with the highest salt intake group. A third analysis used similar
methods.36 In several of the population studies the salt intake in the
lowest salt group was within the usual range of salt intake (6–12 g).
Thus, none of these three analyses provided data on the separate

Table 4
Comparison of original Cochrane review (2003–2017), WASH group Cochrane review (2004–2013) and WHO review (2013): hormones and lipids.

Graudal et al. 2003/2011 (original) MacGregor et al. 2004 (WASH) Aburto et al.
Cochrane 2017 (4) Cochrane 2013 (25) WHO 2013 (27)

N studies 16–88 4–14 4–11

Low sodium, mmol (Hy/No) 80/45 87/78 87/78
Effect renin SMD, (p) 1.22 (0.00001) 0.26 (0.00001) –
Effect aldosterone pg/ml (p) 98 (0.00001) 73 (0.00001) –
Effect noradrenaline pg/ml, (p) 64 (0.00001) 32 (0.01) 8.23 (NS)
Effect adrenalin pg/ml, (p) 8 (0.03) 6.7 (0.06) 6.90 (NS)
Effect cholesterol mg/ml, (p) 5.6 (0.0005) 1.9 (NS) 0.02 (NS)
Effect triglyceride mg/ml, (p) 7 (0.0006) 3.5 (NS) 0.04 (NS)

Hy Hypertensive; No: Normotensive.

25 N. Graudal,
N. Graudal,
G. Jürgens
G. Jürgens
/ Progress
/ Progress
in Cardiovascular
in Cardiovascular
Diseases
Diseases
61 (2018)
61 (2018)
20–26
20–26 25

significance of a low salt intake below 5.8 g. This was done in an IOM re-
port, which concluded that outcomes of population studies were insuf-
ficient to show whether low salt intake below 5.8 g had beneficial or
harmful health effects.37 The first essential meta-analysis to investigate
the separate effect of low salt intake indicated a U-shaped relationship
between salt intake and mortality, especially in study samples represen-
tative for the general population adjusted for multiple effect
modifiers.10 Previously, several individual population studies had iden-
tified this U-shape.31,38–40 Lately, the U-shape was again confirmed in a
meta-analysis of four recent studies.41 A separate analysis of individuals
with HTN based on individual participant data confirmed the U-shape,
but in individuals with normal BP only the low salt intake was associat-
ed with increased mortality, whereas a high salt intake up to 30 g per
day was not.41 Table 5 summarizes the results from the meta-
analyses. All analyses agree that high salt intake above the mean usual
intake is associated with increased mortality in populations of individ-
uals with and without HTN. The latest of the analyses indicate that
this effect only applies to those with HTN. Furthermore the analyses,
which separately investigated low salt intake, agreed that low salt in-
take was associated with increased mortality.10,41
AHA advisory on cohort studies relating salt intake to health outcomes
Representatives of AHA reviewed a series of limitations in the popu-
lation studies, which had the potential to alter the direction of the asso-
ciation between salt intake and health outcomes.12 Potential for
systematic error, for instance in the estimation of salt intake, was iden-
tified in most of the 26 reviewed population studies and specific sys-
tematic error was identified in 6. However, in order to reverse the
direction of the health outcome the systematic error should misclassify
specific groups and such systematic errors were not specifically identi-
fied. Random error due to single estimation of the salt intake was also
potentially present in most studies. Multiple measurements could re-
duce this error and increase the precision of the estimation of the salt
intake,42 which should strengthen the direction of the outcome, but
not reverse the outcome, as verified in two recent studies.32,43 One
study based on multiple 24-h sodium excretions did not find a signifi-
cantly increased (or reduced) rate of CVD44 or all-cause mortality45 in
the low salt group. The authors explained this contrast to the meta-
analyses10,41 as being due to the multiple salt intake estimates used in
these analyses.44,45 However, as it is less likely that multiple measure-
ments reverse the direction of the outcome, the use of less precise
food frequency questionnaires or spot urines to measure the sodium in-
take would probably have yielded similar results. A more reasonable ex-
planation for the lack of U-shape may be that the investigated
individuals suffered from overweight and pre-hypertension or that
few individuals with few events were on a low salt diet, limiting the
power to detect associations in the interval below 5.8 g.
The possibility that sick individuals eat less salt could also explain in-
creased mortality associated with low salt intake (reverse causality).12
We have not seen this hypothetical phenomenon verified in any
study. Conversely, the recently published analysis of NHANES 2014
shows that individuals with HTN, diabetes mellitus, CVD and chronic
kidney disease have salt intakes similar to healthy individuals. This
study also showed that individuals reporting to have intentionally de-
creased their salt intake had the same salt intake as those reporting to
have unchanged salt intake.46 Besides, most of the population studies
reported in the meta-analyses10,41 adjusted for confounders including
diseases. Finally, both of the large meta-analyses10,41 and the largest of
the population studies40 showed that elimination of sick study popula-
tions and sick individuals strengthened the association between low-
salt intake and mortality.
Conclusions
Various biases are prominent in studies supporting salt reduction,
such as selective evaluation of mainly salt sensitive HTN study
populations8,27,34 (Table 1) or salt sensitive overweight study
populations,15,28,44,45 or intentional definition of study inclusion criteria
to increase salt sensitivity of the group of participants being studied,
such as high baseline BP, overweight and reduction of potassium to sub-
normal levels in the control diet,15,44,45 and denial of potential low-salt
side-effects.25,27,34–36 The extraordinary associations of salt with BP and
health outcomes in these studies disappear in subgroup analyses
adjusting for these biases. This selective prioritization in the choice
and methods of evidence to support dietary guidelines, such as Dietary
Guidelines of America (DGA)9 has been criticized previously.47–50 Al-
though the latter50 mainly deals with fat and carbohydrate recommen-
dations, it does emphasize the paradox that DGA says that it “concurs”
with the IOM report, which states that the evidence is “inconsistent
and insufficient to conclude that lowering sodium intakes below 2300
mg/day will have any effect on cardiovascular risk or overall
mortality”37 and yet DGA recommends that sodium intake “should be
less than 2300 mg/day”.9 Recently, this distrust in the process for the es-
tablishment of DGA has been supported in the conclusion of a National
Academy of Medicine report: “Collectively, these findings and conclu-
sions compromise the integrity of the DGA and limit its ability to devel-
op a full body of evidence on a continuous basis over time. The process
to update the DGA should be comprehensively redesigned to allow it to
adapt to changes in needs, evidence, and strategic priorities”.51 Tempo-
rarily, these concerns have had no impact. DGA and other health institu-
tions maintain the idea that the majority of the World's populations
have a too high salt intake. This idea should be evaluated in the context
that this “high” salt intake (6–12 g) is in the low end of the tolerable in-
terval (0.5–55 g), just above the level associated with side effects and in-
creased mortality. A redesign of the salt DGA seems to be needed.
Conflict of interest
None.

Table 5 References
Meta-analyses of population studies: relative risk (RR) for all-cause mortality (ACM), car-
1. McCarron DA, Kazaks AG, Geerling JC, et al. Normal range of human dietary sodium
diovascular disease event (CVD) or stroke.
intake: a perspective based on 24-hour urinary sodium excretion worldwide. Am J
Hypertens. 2013;26:1218-1223.
RR (low salt versus RR (high salt versus
2. Oliver JW, Cohen EL, Neel JV. Blood pressure, sodium intake, and sodium related hor-
usual salt) usual/low salt)
mones in the Yanomamo Indians, a ‘no-salt’ culture. Circulation. 1975;52:146-151.
Reference Populations ACM CVD Stroke ACM CVD Stroke 3. Dahl LK. Possible role of salt intake in the development of essential hypertension. Int J
Epidemiol. 2005;34:967-972.
– – – –
35
Strazzulo All (RPS + IPS) 1.14 1.23a 4. Graudal N, Hubeck-Graudal T, Jürgens G. Effects of sodium restriction on blood pres-
WHO25 All (RPS + IPS) – – – 1.06 1.12 1.24a sure, renin, aldosterone, catecholamines, cholesterols, and triglyceride. Cochrane Da-
Poggio36 All (RPS + IPS) – – – – 1.12a – tabase of Systematic Reviews. 2017(Issue 4). (Art. No: CD004022).
Graudal10 All (RPS + IPS) 1.10a 1.10a 0.96 1.16a 1.12a 1.18a 5. Ofran Y, Lavi D, Opher D, et al. Fatal voluntary salt intake resulting in the highest ever
Graudal10 RPS 1.16⁎ 1.07 0.95 1.04 1.07 1.21a documented sodium plasma level in adults (255 mmol/L): a disorder linked to fe-
Mente41 Normal BP 1.39a 1.28a – 1.00 0.90 – male gender and psychiatric disorders. J Intern Med. 2004;256:525-528.
Mente41 Hypertesion 1.39a 1.35a – 1.39a 1.26a – 6. Geerling JC, Loewy AD. Central regulation of sodium appetite. Exp Physiol. 2008;93:
177-209.
RPS: Representative population samples; IPS: Ill population samples. 7. Brunner HR, Laragh JH, Baer L, et al. Essential hypertension: renin and aldosterone,
a
Statistically significant. heart attack and stroke. N Engl J Med. 1972;286:441-449.
 26 N. Graudal,
N. Graudal,
G. Jürgens
G. Jürgens
/ Progress
/ Progress
in Cardiovascular
in Cardiovascular
8. Institute of Medicine. Dietary Reference Intakes: Water, Potassium, Sodium, Chloride,
and Sulfate. Washington, DC: National Academies Press. 2005:379-380.
9. Dietary Guidelines for the Americans. , https://health.gov/dietaryguidelines/2015/
guidelines/ 2015-2020.
10. Graudal N, Jürgens G, Baslund B, Alderman MH. Compared with usual sodium intake,
low- and excessive-sodium diets are associated with increased mortality: a
meta-analysis. Am J Hypertens. 2014;27:1129-1137.
11. Cogswell ME, Mugavero K, Bowman BA, et al. Dietary sodium and cardiovascular dis-
ease risk - measurement matters. N Engl J Med. 2016;375:580-586.
12. Cobb LK, Anderson CA, Elliott P, et al. Methodological issues in cohort studies that re-
late sodium intake to cardiovascular disease outcomes: a science advisory from the
American Heart Association. Circulation. 2014;129:1173-1186.
13. Johnson AG, Nguyen TV, Davis D. Blood pressure is linked to salt intake and modulat-
ed by the angiotensinogen gene in normotensive and hypertensive elderly subjects. J
Hypertens. 2001;19:1053-1060.
14. MacGregor GA, Markandu ND, Sagnella GA, et al. Double-blind study of three sodium
intakes and long-term effects of sodium restriction in essential hypertension. Lancet.
1989;2:1244-1247.
15. Sacks FM, Svetkey LP, Vollmer WM, et al. Effects on blood pressure of reduced dietary
sodium and the dietary approaches to stop hypertension (DASH) diet. N Engl J Med.
2001;344:3-10.
16. Bruun NE, Skøtt P, Nielsen MD, et al. Normal renal tubular response to changes of so-
dium intake in hypertensive man. J Hypertens. 1990;8:219-227.
17. Graudal NA, Hubeck-Graudal T, Jurgens G, McCarron DA. The significance of duration
and dose of sodium reduction intervention in normotensive and hypertensive indi-
viduals. A meta-analysis. Adv Nutr. 2015;6:169-177.
18. Appel LJ, Moore TJ, Obarzanek E, et al. A clinical trial of the effects of dietary patterns
on blood pressure. DASH collaborative research group. N Engl J Med. 1997;336:
1117-1124.
19. Cappuccio FP, MacGregor GA. Does potassium supplementation lower blood pres-
sure? A meta-analysis of published trials. J Hypertens. 1991;9:465-473.
20. Whelton PK, He J, Cutler JA, et al. Effects of oral potassium on blood pressure.
Meta-analysis of randomized controlled clinical trials. JAMA. 1997;277:1624-1632.
21. Sacks FM, Campos H. Dietary therapy in hypertension. N Engl J Med. 2010;362:
2102-2112.
22. Kaiser J. Industry groups petition for data on salt and hypertension. Science.
2003;300:1350.
23. He FJ, Li J, MacGregor GA. Effect of longer-term modest salt reduction on blood pres-
sure. Cochrane Database Syst Rev. 2013(Issue 4). (Art. No: CD004937).
24. WHO. Sodium Intake for Adults and Children. Geneva, Switzerland: WHO. 2012.
25. Aburto NJ, Ziolkovska A, Hooper L, et al. Effect of lower sodium intake on health: sys-
tematic review and meta-analyses. BMJ. 2013;346:f1326.
26. Frieden TR. Sodium reduction—saving lives by putting choice into consumer's hands.
JAMA. 2016;316:579-580.
27. Mozaffarian D, Fahimi S, Singh GM, et al. Global sodium consumption and death from
cardiovascular causes. N Engl J Med. 2014;37:624-634.
28. Jackson SL, Cogswell ME, Zhao L, et al. Association between urinary sodium and po-
tassium excretion and blood pressure among adults in the United States: National
Health and Nutrition Examination Survey, 2014. Circulation. 2018;137:237-246.
29. Mente A, O'Donnell MJ, Rangarajan S, et al. Association of urinary sodium and potas-
sium excretion with blood pressure. N Engl J Med. 2014;371:601-611.
abstrak Sembilan puluh lima persen populasi Dunia
memiliki asupan garam rata-rata antara 6 dan 12 g, yang
jauh lebih rendah daripada tingkat harian yang ditoleransi
hingga 55 g / d. Meskipun demikian, tingkat atas yang
direkomendasikan oleh banyak lembaga kesehatan
adalah serendah 5,8 g / hari. Ketika meninjau bukti untuk
tingkat atas 5,8 g / hari, menjadi jelas bahwa baik studi
pendukung yang dipilih oleh lembaga kesehatan, atau uji
coba terkontrol secara acak dan studi observasional
prospektif diabaikan oleh institusi kesehatan,
mendokumentasikan bahwa asupan garam di bawah ini
5,8 g ini, memiliki efek kesehatan yang bermanfaat.
Meskipun ada hubungan antara asupan garam dan
tekanan darah, baik dalam uji coba terkontrol acak dan
dalam penelitian observasional, hubungan ini lemah,
terutama pada individu yang tidak obesitas dengan
tekanan darah normal. Selanjutnya asupan garam di
bawah 5,8 g dikaitkan dengan aktivasi sistem renin-
angiotensin-aldosteron, peningkatan lipid plasma dan
peningkatan mortalitas. Desain ulang dari pedoman diet
garam, oleh karena itu, tampaknya diperlukan
pengantar Sembilan puluh lima persen populasi dunia
memiliki asupan garam rata-rata antara 6 dan 12 g / d, 1
Diseases
Diseases
61 (2018)
61 (2018)
20–26
20–26 26
jumlah minimum yang diperlukan adalah sekitar 0,5 g /
d.2 Asupan garam hingga sekitar 55 g / d telah dicatat.3
Percobaan terkontrol acak (RCT) dari asupan garam
pada interval 0,5–40 g / d belum melaporkan defisiensi
atau gejala toksik. Intoksikasi telah dijelaskan setelah
asupan cepat sekitar 50 g atau lebih selama beberapa
menit.5 Garam sangat penting untuk kehidupan, karena
berkontribusi terhadap potensi aksi dan potensi membran
sel dan mempertahankan volume ekstraseluler dan
tekanan darah (BP) .6 Pusat di otak mengatur garam
tubuh, bersama dengan sistem renin-aldosterone-
angiotensin (RAAS) dan ginjal.6,7 Namun, banyak
lembaga kesehatan setuju bahwa garam sama
beracunnya dengan tembakau dan, oleh karena itu,
menganggap garam sebagai target untuk
pencegahan.8,9 Posisi ini didasarkan pada keyakinan
bahwa garam tidak hanya mempertahankan BP, tetapi
juga meningkatkan BP sebagai fungsi linear dari jumlah
tertelan, yang mengarah ke peningkatan mortalitas.8
Dalam beberapa tahun terakhir sejumlah besar RCT dan
studi populasi telah mempertanyakan efek berbahaya dari
garam. 4,10 Banyak institusi kesehatan tidak setuju dan
masih mendukung intervensi untuk mengurangi asupan
garam pada populasi umum hingga di bawah 5,8 g8,9,11
secara paralel. dengan upaya untuk menolak hasil
penelitian yang menunjukkan efek berbahaya dari asupan
garam yang rendah.12 Repre-sentatif dari American
Heart Association (AHA) meninjau kualitas metodologis
dari 26 studi populasi untuk menganalisis apakah
masalah metodologis menyumbang kurangnya efek
menguntungkan dari rendah asupan garam. Mereka
menyimpulkan bahwa penelitian ini, karena masalah
metodologis, tidak cocok untuk membentuk dasar
pedoman garam, yang seharusnya didasarkan pada
“bukti kuat yang menghubungkan Na dengan tekanan
darah tinggi dan beberapa uji populasi umum yang ada
dari efek pengurangan Na pada penyakit kardiovaskular
(CVD) ”.12 Namun, sebagaimana ditunjukkan dalam
meta-analisis RCT4, bukti ini mungkin tidak sekuat yang
sebelumnya diklaim. 12 Oleh karena itu kami menemukan
hal yang relevan untuk menyajikan tinjauan kritis kolektif
terhadap kualitas bukti kesehatan utama untuk
pengurangan garam (SR) yang dipromosikan oleh
institusi kesehatan dalam konteks bukti yang ada dari
RCT dan studi populasi prospektif.
Metode
Kami memasukkan RCT dan studi populasi serta penelitian
berdasarkan RCT dan studi populasi (meta-analisis dan studi
pemodelan) yang dimulai atau didukung secara ekonomi oleh
lembaga dan organisasi kesehatan berikut, yang memberikan
prioritas tinggi pada pengurangan garam: Pusat Pengendalian
Penyakit dan Pencegahan (CDC), Institut Kesehatan Nasional
(NIH), National Heart, Lung and Blood Institute (NHLBI),
Institute of Medicine (IOM, sekarang National Academy of
Medicine (NAM)), American Heart Association (AHA),
Makanan dan Drug Association (FDA), World Health
Organization (WHO), dan World Action on Salt and Health
(WASH). Identifikasi penelitian sebagian didasarkan pada
pencarian sistematis kami sebelumnya RCT4 dan studi
populasi, 10 dan daftar referensi dari laporan, studi dan
 27 N. Graudal,
N. Graudal,
G. Jürgens
G. Jürgens
/ Progress
/ Progress
in Cardiovascular
in Cardiovascular
makalah posisi yang diterbitkan oleh lembaga-lembaga yang
disebutkan.
Garam: berlaku pada BP
IOM (sekarang NAM) melaporkan asupan referensi garam diet
Batas atas untuk asupan garam 5,8 g yang direkomendasikan
oleh lembaga kesehatan berasal dari laporan IOM, 8 yang
menyatakan, “paling relevan untuk menentukan tingkat atas
adalah tiga percobaan di mana tingkat terendah asupan natrium
diet dekat dengan asupan yang memadai (Johnson et al., 2001;
MacGregor et al., 1989; Sacks et al., 2001). ”“ Mengingat hasil
dari ketiga percobaan ini, tingkat efek samping terendah yang
diamati untuk diet natrium diatur pada 2,3 g / hari (5,8 g garam
/ hari). ”3 penelitian ini, 13-15 individu yang diacak untuk 3-4
dosis garam (analisis respons dosis), semuanya menunjukkan
proporsi yang signifikan
peningkatan BP dengan peningkatan asupan garam di bawah
dan di atas asupan garam 5,8 g, sehingga membenarkan
kesimpulan bahwa asupan garam harus di bawah 5,8 g.
Gambar. 1 menunjukkan semua studi dosis-respons yang
diketahui yang diterbitkan pada saat evaluasi IOM. Tiga
penelitian13-15 yang dipertimbangkan oleh IOM adalah semua
penelitian pada individu yang lebih tua dengan hipertensi
(HTN) dan mereka dengan hubungan dosis-respons yang paling
tajam. Penelitian HTN dengan hubungan dosis-respons
terendah16 dan empat studi normoten- sive tanpa hubungan
dosis-respons (Gambar 1) 16,17 tidak dimasukkan dalam
evaluasi. Inklusi selektif ini lebih lanjut dikukuhkan oleh fakta
bahwa BP tinggi dan usia bias tiga studi yang dipilih.
Studi pertama termasuk individu yang lebih tua dengan sangat
tinggi BP.13 Mem-
Anggota kelompok WASH melakukan penelitian kedua.14
Kelompok ini telah menerbitkan 10 penelitian pada individu
dengan HTN yang menunjukkan efek rata-rata SR pada sistolik
BP (SBP) sekitar 10 mm Hg, yang dua kali lipat dari yang
dilaporkan dalam semua RCT lain dari HTN. (4,85 mm Hg)
(Tabel 1). Ini menunjukkan bias sistematis, yang hanya
sebagian tergantung pada BP dasar yang tinggi dari populasi
yang termasuk. Ketua komite IOM tahun 20057 adalah penulis
dari studi ketiga. Penelitian ini, studi Dietary Approaches to
Stop Hypertension (DASH), adalah yang paling disorot dari
penelitian yang digunakan untuk membenarkan SR.
Pendekatan Diet untuk Menghentikan Hipertensi (DASH)
studies15,18 Dalam studi DASH pertama, 3 diet
dibandingkan, diet kontrol selesai dalam potasium,
kalsium dan magnesium ke 25% persentil dari populasi
dalam rangka “untuk memastikan kontras yang jelas
dengan pola diet yang ideal”. Diet ideal kaya kalium,
kalsium, magnesium, karbohidrat, protein dan serat dan
rendah lemak. Diet ketiga adalah di antara dua diet ini.
Pada saat desain penelitian ini diketahui bahwa asupan
kalium berbanding terbalik dengan BP.19,20 Dengan
demikian efek reduksi BP dari buah kaya kalium yang
ditambahkan pada diet DASH, yang diidentifikasi dalam
penelitian ini, mungkin sebagian dianggap berasal dari
deplesi potassium yang dirancang dalam kelompok
kontrol. Pengalaman dari sidang DASH18 pertama
dilakukan ke depan uji coba DASH sodium berikutnya, 15
Diseases
Diseases
61 (2018)
61 (2018)
20–26
20–26 27
di mana diet menengah dihilangkan dan para partisipan
diacak untuk ideal dan
kontrol diet, masing-masing, dan selain itu menyeberang ke tiga
diet asupan garam yang berbeda. SR mengurangi SBP sebesar
6,7 mm Hg dalam diet kontrol kalium yang ditentukan dan
dengan 3 mm Hg dalam diet yang ideal. Sebuah meta-analisis
dari lima penelitian yang diidentifikasi di antara 176 penelitian
pengurangan garam, 4 yang selain alokasi untuk diet rendah
garam juga mengalokasikan peserta untuk kombinasi
pengurangan natrium dan suplementasi kalium, menunjukkan
perbedaan TD yang signifikan antara natrium rendah. / diet
kalium rendah dibandingkan dengan diet rendah sodium / tinggi
kalium (Gbr. 2). Perbedaan ini sesuai dengan perbedaan yang
diamati dalam uji coba DASH natrium, yang menunjukkan
bahwa pengurangan kalium yang direncanakan pada kelompok
kontrol meliputi penurunan BP yang disebabkan oleh
pengurangan garam. Lebih jauh lagi, usia rata-rata, indeks
massa tubuh (BMI) dan baseline BP dari peserta yang
diikutsertakan secara signifikan lebih tinggi daripada rata-rata
penduduk Amerika. Bias umum ini tercermin dalam publikasi
tambahan dari percobaan DASH, yang menunjukkan bahwa
pada individu yang lebih muda antara 21 dan 42 tahun,
mewakili lebih dari 50% populasi Amerika, diet rendah sodium
/ tinggi yang disediakan tidak berpengaruh pada SBP.
dibandingkan dengan diet rendah sodium / rendah yang habis.
Karena desain, interpretasi dan penggunaan uji coba DASH-
sodium, itu adalah batasan utama bahwa data dari uji coba yang
didanai pemerintah ini tidak tersedia untuk umum.
WASH dan WHO versus Cochrane: meta-analisis dari efek
pengurangan asupan garam pada BP
Cochrane telah menerbitkan dua ulasan reduksi garam oleh dua
kelompok penulis yang berbeda yang mengukur hasil yang
sama persis4,23; justifikasinya adalah bahwa ulasan asli4
menginvestigasi efek akut SR, sedangkan tinjauan terbaru dari
2004 oleh anggota WASH
group23 menyelidiki efek jangka panjang. Namun, pada
tahun 2004 tidak ada ilmiah pembenaran untuk
perbedaan ini dan review kemudian longi- tudinal RCT
menunjukkan tidak ada perbedaan efek SR pada BP
antara minggu 1 dan minggu 6.17 The review24,25 WHO
meliputi hampir studi sama WASH yang ulasan. Tabel 2
membandingkan tinjauan Cochrane asli, tinjauan WASH
dan tinjauan WHO. Secara umum tidak ada perbedaan
antara efek BP yang memverifikasi bahwa perbedaan
antara studi akut dan jangka panjang tidak dibenarkan.
Efek sedikit lebih tinggi dalam analisis studi normotensif
dalam tinjauan WASH adalah karena efek studi DASH
dan 3 penelitian, yang termasuk individu normotensif dan
mereka dengan HTN (Tabel 3). Setelah pengecualian dari
4 penelitian ini 8 studi yang tersisa menunjukkan efek,
yang hampir identik dengan ulasan Cochrane dan ulasan
WHO (Tabel 2). Bukti CDC / FDA untuk hubungan antara
asupan garam diet dikurangi dan BP Baru-baru ini, CDC
dan FDA merilis proposal untuk panduan sukarela untuk
mendorong perusahaan makanan untuk mengurangi
natrium dalam makanan dan makanan restoran.11,26
Argumennya adalah “bukti kuat, termasuk analisis terbaru
 28 N. Graudal,
N. Graudal,
G. Jürgens
G. Jürgens
/ Progress
/ Progress
in Cardiovascular
in Cardiovascular
terhadap lebih dari 100 penelitian acak. uji coba, bahwa
pengurangan natrium mengurangi tekanan darah pada
orang dewasa. ”Namun, analisis ini27 didasarkan pada
65% studi HTN dan 35% penelitian normotensif. Garis
meta-regresi dengan kemiringan 3,8 mm Hg per 100
mmol sodium (2,3 g sodium) dipaksa melalui nol dan
terutama berdasarkan data yang diadopsi dari tinjauan
Cochrane yang asli. 4 Fungsi yang sesuai dengan
konstanta mengungkapkan bahwa kemiringan hanya 2,27
mm Hg / 100 mmol (Gbr. 3). Para penulis menerapkan
linearitas tanpa-konstan dari analisis meta-regresi
campuran baik ke HTN maupun
Dalam artikel utama analisis terbaru ini menunjukkan hubungan
antara asupan natrium dan BP, yang lebih kuat daripada yang
ditemukan dalam studi populasi sebelumnya (4,58 / 2,25 mm
Hg / 1 g Na) .28 Namun, menurut eTable 2 dalam suplemen
mereka, efek ini terutama karena adiposa 50% dari populasi
dengan BMI di atas 30. Dalam kelompok peserta dengan BMI
kurang dari 30 efek tekanan darah sistolik hanya 1,8 mm Hg / 1
g Na. Hal ini mirip dengan efek yang ditemukan dalam studi
prospektif Perkotaan Pedesaan Epistemologi (MURNI) di
seluruh dunia (2,1 mm Hg / 1 g Na), di mana populasi
penelitian memiliki BMI rata-rata 26,29
Garam: efek pada hormon dan lipid
WASH dan WHO versus Cochrane: meta-analisis dari efek
pengurangan asupan garam pada hormon dan lipid
Efek dari pengurangan garam pada renin, aldosteron,
noradrenalin, adrenalin, kolesterol dan trigliserida telah
dianalisis dalam tiga meta-analisis yang telah disebutkan
sebelumnya. Hasil ini disajikan pada Tabel 4. Ulasan Cochrane
yang asli4 mengurangi garam. asupan ke a
tingkat rata-rata di bawah 5,0 g sesuai dengan rekomendasi,
sedangkan dua analisis kecil studi minimal 4 minggu
duration23,25 mengurangi asupan garam ke, tetapi tidak di
bawah, tingkat rata-rata 5,0 g sesuai dengan rekomendasi
WHO, yang untuk mengurangi garam di bawah 5.0 g / d.24
Dengan demikian, tingkat asupan garam, daripada durasi
paparan, menentukan terjadinya efek samping. Ini telah
diverifikasi di Indian Yanomamo pada asupan rendah garam,
yang terus-menerus meningkatkan kadar renin dan aldosteron
dalam darah2; dan dalam studi cross-sectional individu dengan
BP normal dan hipertensi. RAAS diaktifkan oleh pengurangan
garam pada individu yang sehat dan sakit, tetapi McCarron
menyarankan bahwa pada pasien dengan gagal jantung (HF)
dan penyakit ginjal, yang RAAS diaktifkan oleh perfusi ginjal
yang dikompromikan, RAAS yang aktif mungkin mendiktekan
peningkatan yang didorong saraf dalam asupan garam dalam
upaya untuk meningkatkan perfusi ginjal dan menekan aktivitas
renin plasma dan angiotensin II dan dampak patologisnya pada
jantung dan pembuluh darah. keadaan ini asupan garam yang
tinggi disarankan untuk tidak menjadi penyebab, tetapi lebih
mungkin mekanisme kompensasi kompensasi, berfungsi
sebagai inhibitor RAAS alami. 30 Interpretasi tersebut
konsisten dengan penelitian terbaru tentang pasien dengan
gagal jantung dan penyakit ginjal yang menunjukkan bahwa
garam rendah. diet tidak terkait dengan penurunan morbiditas
atau mortalitas.31,32
Diseases
Diseases
61 (2018)
61 (2018)
20–26
20–26 28
Garam: efek pada hasil kesehatan
RCT berkaitan dengan asupan garam untuk hasil kesehatan
RCT mengukur efek kesehatan telah dilakukan pada individu
dengan HTN dan individu kelebihan berat badan pra-HTN,
tetapi tidak sehat
individu. Secara bersama-sama mereka menunjukkan
pengurangan 24% yang tidak signifikan pada kejadian
CVD pada kelompok rendah garam (data yang tidak
tercatat secara lengkap) dan tidak ada perbedaan dalam
semua penyebab kematian (data tercatat secara lengkap)
.33 Asupan garam yang berarti pada kelompok rendah
garam adalah 5,8 g atau lebih tinggi. Dengan demikian
tidak ada RCT untuk menunjukkan efek kesehatan
asupan garam di bawah 5,8 g. Studi pemodelan yang
menghubungkan asupan garam dengan hasil kesehatan
Studi pemodelan menetapkan hubungan dosis-respons
antara asupan garam dan BP, yang secara tidak
langsung digunakan untuk menerjemahkan pengurangan
garam ke pengurangan kematian dengan menggunakan
data dari studi observasional yang menghubungkan BP
dengan mortalitas. Satu studi pemodelan34
menggunakan analisis regresi linier berdasarkan data dari
meta-analisis WASH (Tabel 3), data dari studi DASH15
(Gambar 1) dan data dari salah satu studi kelompok
WASH14 (Gambar 1) (studi no. 4 pada Tabel 1) untuk
membangun analisis dosis-respons. Penelitian lain27
menggunakan data dari analisis meta-regresi yang
dijelaskan di atas 103 RCT, dimana 65% adalah HTN.
Tak satu pun dari studi pemodelan termasuk data efek
samping dalam model, meskipun data efek samping
tersedia dalam meta-analisis dari mana data BP diadopsi.
Model-model ini meramalkan ribuan hingga jutaan jiwa
yang diselamatkan oleh pengurangan garam makanan,
berbeda dengan data nyata dari penelitian kohort, yang
menunjukkan bahwa asupan garam yang rendah
dikaitkan dengan peningkatan mortalitas. Studi kohort
yang menghubungkan asupan garam dengan hasil
kesehatan Bukti dari WHO didasarkan pada meta-analisis
dari 2009,35 yang diperbarui pada tahun 2013.25 Pada
analisis 2009, risiko relatif dari asupan garam yang lebih
tinggi dan lebih rendah diselidiki dengan membandingkan
tingkat kejadian dalam dua kategori dengan perbedaan
rata-rata konsumsi garam paling dekat dengan 5,8 g /
hari. Dalam analisis yang diperbarui, estimasi efek
keseluruhan adalah gen erated membandingkan risiko
setiap hasil dalam kelompok asupan garam terendah
dengan kelompok asupan garam tertinggi. Analisis ketiga
menggunakan metode serupa.36 Dalam beberapa studi
populasi, asupan garam pada kelompok garam terendah
berada dalam kisaran asupan garam biasa (6-12 g).
Dengan demikian, tidak satupun dari ketiga analisis ini
menyediakan data terpisah
signifikansi asupan garam rendah di bawah 5,8 g. Hal ini
dilakukan dalam laporan IOM, yang menyimpulkan bahwa
hasil penelitian populasi tidak cukup untuk menunjukkan
apakah asupan garam yang rendah di bawah 5,8 g
memiliki efek kesehatan yang menguntungkan atau
berbahaya.37 Meta-analisis esensial pertama untuk
 29 N. Graudal,
N. Graudal,
G. Jürgens
G. Jürgens
/ Progress
/ Progress
in Cardiovascular
in Cardiovascular
menyelidiki efek terpisah dari Asupan garam rendah
menunjukkan hubungan berbentuk U antara asupan
garam dan mortalitas, terutama dalam sampel penelitian
yang representatif untuk populasi umum yang
disesuaikan untuk beberapa pengubah efek.10
Sebelumnya, beberapa studi populasi individu telah
diidentifikasi. dengan bentuk U ini.31,38–40 Akhir-akhir
ini, bentuk U kembali dikonfirmasi dalam a meta-analisis
dari empat penelitian terbaru. Sebuah analisis terpisah
dari individu dengan HTN berdasarkan data individu
peserta dikonfirmasi bentuk U, tetapi pada individu
dengan BP normal hanya asupan garam rendah dikaitkan
dengan peningkatan mortalitas, sedangkan garam tinggi.
Asupan hingga 30 g per hari tidak.41 Tabel 5 merangkum
hasil dari meta- analisis. Semua analisis setuju bahwa
asupan garam yang tinggi di atas rata-rata asupan biasa
dikaitkan dengan peningkatan mortalitas pada populasi
individu dengan dan tanpa HTN. Yang terbaru dari
analisis menunjukkan bahwa efek ini hanya berlaku untuk
mereka yang memiliki HTN. Selanjutnya analisis, yang
secara terpisah menyelidiki asupan rendah garam, setuju
bahwa asupan garam yang rendah dikaitkan dengan
peningkatan mortalitas.10,41
Penasihat AHA pada penelitian kohort yang
menghubungkan asupan garam dengan hasil kesehatan
Perwakilan dari AHA meninjau serangkaian keterbatasan
dalam studi populasi, yang memiliki potensi untuk
mengubah arah asosiasi antara asupan garam dan hasil
kesehatan. Potensi untuk kesalahan sistematis, misalnya
dalam estimasi asupan garam, diidentifikasi di sebagian
besar dari 26 penelitian populasi dan kesalahan sistem
spesifik diidentifikasi pada 6. Namun, untuk membalikkan
arah hasil kesehatan kesalahan sistematis harus
misclassify kelompok tertentu dan kesalahan sistematis
seperti itu tidak secara khusus identi fi - disaring.
Kesalahan acak karena estimasi tunggal dari asupan
garam juga berpotensi hadir dalam sebagian besar
penelitian. Pengukuran berganda dapat mengurangi
kesalahan ini dan meningkatkan ketepatan estimasi
asupan garam, 42 yang seharusnya memperkuat arah
hasil, tetapi tidak membalik hasilnya, sebagaimana
diverifikasi dalam dua penelitian terbaru.32,43 Satu
penelitian berdasarkan ekskresi natrium 24 jam tidak
menemukan peningkatan signifikan (atau penurunan)
tingkat CVD44 atau semua penyebab mortalitas45 pada
kelompok rendah garam. Para penulis menjelaskan
kontras ini dengan meta-analisis10,41 karena perkiraan
perkiraan asupan garam yang digunakan dalam analisis
ini.44,45 Namun, karena kurang mungkin bahwa
beberapa pengukuran membalik arah hasil, penggunaan
kuesioner frekuensi makanan yang kurang tepat atau urin
spot untuk mengukur asupan natrium mungkin akan
menghasilkan hasil yang sama. Penjelasan yang lebih
masuk akal untuk kurangnya bentuk-U mungkin adalah
bahwa individu yang diteliti menderita kelebihan berat
badan dan pra-hipertensi atau bahwa beberapa individu
dengan beberapa kejadian berada pada diet rendah
garam, membatasi kekuatan untuk mendeteksi asosiasi
dalam interval di bawah ini. 5,8 g. Kemungkinan bahwa
Diseases
Diseases
61 (2018)
61 (2018)
20–26
20–26 29
orang yang sakit memakan lebih sedikit garam juga dapat
menjelaskan peningkatan mortalitas yang terkait dengan
asupan garam yang rendah (kausalitas terbalik) .12
Kami belum melihat fenomena hipotetis ini diverifikasi dalam
penelitian apa pun. Sebaliknya, analisis NHANES 2014 yang
baru-baru ini dipublikasikan menunjukkan bahwa individu
dengan HTN, diabetes mellitus, CVD dan penyakit ginjal kronis
memiliki asupan garam yang mirip dengan individu yang sehat.
Penelitian ini juga menunjukkan bahwa orang yang melaporkan
untuk secara sengaja mengurangi asupan garam mereka
memiliki asupan garam yang sama seperti mereka yang
melaporkan memiliki asupan garam tidak berubah. Selain itu,
sebagian besar studi populasi yang dilaporkan dalam meta-
analisis10,41 disesuaikan untuk pembaur termasuk penyakit
Akhirnya, kedua meta-analisis besar10,41 dan studi populasi
terbesar40 menunjukkan bahwa eliminasi populasi penelitian
sakit dan individu yang sakit memperkuat hubungan antara
asupan garam rendah dan mortalitas
Kesimpulan Berbagai bias menonjol dalam penelitian
yang mendukung pengurangan garam, seperti evaluasi
selektif terutama populasi studi HTN sensitif
garam8,27,34 (Tabel 1) atau populasi studi overweight
garam sensitif, 15,28,44,45 atau definisi disengaja kriteria
inklusi studi untuk meningkatkan sensitivitas garam dari
kelompok peserta yang diteliti, seperti BP awal yang
tinggi, kelebihan berat badan dan pengurangan kalium ke
tingkat sub-normal dalam diet kontrol, 15,44,45 dan
penolakan potensi efek samping rendah garam. , 27,34-
36 Asosiasi garam yang luar biasa dengan BP dan hasil
kesehatan dalam studi ini menghilang dalam analisis
subkelompok yang disesuaikan untuk bias ini. Ini prioritas
selektif dalam pilihan dan metode bukti untuk mendukung
pedoman diet, seperti Dietary Guidelines of America
(DGA) 9 telah dikritik sebelumnya. 47-50 Meskipun yang
terakhir50 terutama berkaitan dengan rekomendasi lemak
dan karbohidrat, itu tidak menekankan paradoks bahwa
DGA mengatakan bahwa “sependapat” dengan laporan
IOM, yang menyatakan bahwa bukti “tidak konsisten dan
tidak cukup untuk menyimpulkan bahwa menurunkan
asupan natrium di bawah 2300 mg / hari akan memiliki
efek pada risiko kardiovaskular atau mortalitas
keseluruhan” 37 dan namun DGA merekomendasikan
bahwa asupan natrium “harus kurang dari 2.300 mg /
hari” .9 Baru-baru ini, ketidakpercayaan ini dalam proses
untuk pembentukan DGA telah didukung dalam
kesimpulan laporan National Academy of Medicine:
“Secara kolektif, ini temuan dan kesimpulan
mengkompromikan integritas DGA dan membatasi
kemampuannya untuk mengembangkan seluruh bukti
yang terus menerus dari waktu ke waktu. Proses untuk
memperbarui DGA harus dirancang ulang secara
komprehensif agar memungkinkan untuk beradaptasi
dengan perubahan kebutuhan, bukti, dan prioritas
strategis ”.51 Sementara itu, kekhawatiran ini tidak
berdampak. DGA dan lembaga kesehatan lainnya
mempertahankan gagasan bahwa mayoritas populasi
Dunia memiliki asupan garam yang terlalu tinggi. Ide ini
harus dievaluasi dalam konteks bahwa asupan garam
“tinggi” ini (6-12 g) berada di ujung rendah dari interval
yang dapat ditoleransi (0,5–55 g), tepat di atas tingkat
 30 N. Graudal,
N. Graudal,
G. Jürgens
G. Jürgens
/ Progress
/ Progress
in Cardiovascular
in Cardiovascular
Diseases
Diseases
61 (2018)
61 (2018)
20–26
20–26 30

yang terkait dengan efek samping dan efek samping.
kematian berkerut. Desain ulang dari DGA garam
tampaknya diperlukan.
30. McCarron DA. What determines human sodium intake: policy or physiology? Adv
Nutr. 2014;5:578-584.
31. O'Donnell MJ, Yusuf S, Mente A, et al. Urinary sodium and potassium excretion and
risk of cardiovascular events. JAMA. 2011;306:2229-2238.
32. Fan L, Tighiouart H, Levey AS, et al. Urinary sodium excretion and kidney failure in
nondiabetic chronic kidney disease. Kidney Int. 2014;86:582-588.
33. Adler AJ, Taylor F, Martin N, et al. Reduced dietary salt for the prevention of cardio-
vascular disease. Cochrane Database Syst Rev. 2014;12, CD009217.
34. Bibbins-Domingo K, Chertow GM, Coxson PG, et al. Projected effect of dietary salt re-
ductions on future cardiovascular disease. N Engl J Med. 2010;362:590-599.
35. Strazzullo P, D'Elia L, Kandala NB, Cappuccio FP. Salt intake, stroke, and cardiovascular
disease: meta-analysis of prospective studies. BMJ. 2009;339:b4567.
36. Poggio R, Gutierrez L, Matta MG, et al. Daily sodium consumption and CVD mortality
in the general population: systematic review and meta-analysis of prospective stud-
ies. Public Health Nutr. 2015;18:695-704.
37. IOM. (Institute of Medicine). Sodium Intake in Populations: Assessment of Evidence.
Washington, DC: National Academies Press. 2013.
38. Thomas MC, Moran J, Forsblom C, et al. The association between dietary sodium intake, ESRD,
and all-cause mortality inpatients with type 1 diabetes. Diabetes Care. 2011;34:861-866.
39. Phister R, Michels G, Sharp SJ, et al. Estimated urinary sodium excretion and risk of
heart failure in men and women in the EPIC-Norfolk study. Eur J Heart Fail.
2014;16:394-402.
40. O'Donnell M, Mente A, Rangarajan S, et al. Urinary sodium and potassium excretion,
mortality, and cardiovascular events. N Engl J Med. 2014;371:612-623.
41. Mente A, O'Donnell M, Rangarajan S, et al. Associations of urinary sodium ex-
cretion with cardiovascular events in individuals with and without hyperten-
sion: a pooled analysis o f data from four studies. Lancet. 2016;388(10043):
465-475.
42. Lerchl K, Rakova N, Dahlmann A, et al. Agreement between 24-hour salt inges-
tion and sodium excreti on in a cont rolled environm ent. Hypertension.
2015;66:850-857.
43. Olde Engberink RHG, van den Hoek TC, van Noordenne ND, et al. Use of a single
baseline versus multiyear 24-hour urine collection for estimation of long-term
sodium int ak e and a s sociated cardi ovas cular a nd ren al risk . Circul ation.
2017;136:917-926.
44. Cook NR, Appel LJ, Whelton PK. Lower levels of sodium intake and reduced cardio-
vascular risk. Circulation. 2014;129:981-989.
45. Cook NR, Appel LJ, Whelton PK. Sodium intake and all-cause mortality over 20 years
in the trials of hypertension prevention. J Am Coll Cardiol. 2016;68:1609-1617.
46. Cogswell ME, Loria CM, Terry AL, et al. Estimated 24-hour urinary sodium and potas-
sium excretion in US adults. JAMA (doi:10.1001/jama.2018.1156).
47. Folkow B. On bias in medical research; reflections on present salt- cholesterol contro-
versies. Scand Cardiovasc J. 2011;45:194-197.
48. Graudal N. A radical sodium reduction policy is not supported by randomized controlled
trials or observational studies: grading the evidence. Am J Hypertens. 2016;29:543-548.
49. Graudal N. Reducing salt intake at the population level: is it really a public health
priority? Nephrol Dial Transplant. 2016;31:1398-1403.
50. Teicholz N. The scientific report guiding the US dietary guidelines: is it scientific?
BMJ. 2015;351:h4962.
51. National Academy of Medicine. Redesigning the Process for Establishing the Dietary
Guidelines for Americans. Washington: National Academy of Sciences. 2017.