Periodically studies are mentioned in the popular press from the medical literature saying you don’t need to reduce the sodium in your diet. They try to convey the idea that you can take in any amount of sodium and it will make no difference to your health. This is based on a relatively small number of studies. And the studies have a lot of problems. In 2012 (1) a study came out analyzing some recent studies supporting the idea that a low sodium diet can be dangerous.
There are two relatively recent studies that claim getting sodium below 1500 to 3000 mg is dangerous and makes you more likely to die. These two studies describe what has become a popular concept concerning the danger of a low sodium diet. They claim there is a J-shaped or U-shaped relationship between sodium intake and serious health problems.
Both studies have serious design problems. The studies were done in patients who were at high risk of stroke, cardiovascular disease (CVD), kidney disease and other disease. The studies determined dietary sodium by collecting a 24 hour urine sample for sodium in the first study, and a single morning spot urine sample for sodium in the second study. Such sampling methods are not as accurate as methods using multiple collections, since day to day intake of sodium varies a lot.
The first study (2) was a 10 year follow-up of over 2800 Finnish adults. The researchers found a J-shaped survival curve in relation to sodium. From their graph, survival improved as daily sodium was lowered until it reached somewhere between 1200 and 1700 mg per day. Survival then worsened at lower sodium levels, implying danger for a low sodium diet.
For 75% of the participants the rate of death from all causes increased proportionally to sodium. For the 25% taking in a lower amount of sodium an uptick in deaths occurred. The researchers did not report separately on cardiovascular events and deaths, or stroke events and deaths.
Problems With First Study
The biggest problem with the study is that it was in high risk patients. This is done in many medical studies so enough adverse events can occur to make the findings significant. This is a valid method for determining how to medically manage high risk patients. It is not valid for studies of prevention. The object of prevention studies is to find what can allow people to avoid becoming a high risk patient.
The participants at the time of entry into the study already had diabetes for a median of 20 years duration. 47% had hypertension, and 30% were spilling albumin (protein) in their urine.
The protein in the urine indicates some degree of kidney disease. The low sodium group had more participants with very poor kidney function. The high sodium group had some participants with kidney disease, but kidney function was better than in the low sodium group.
In this study, as the sodium in the urine (assumed in the proper situation to be proportional to how much is consumed) went down, the function of the participants’ kidneys was worse. This means that many of the participants already had damaged kidneys. The low sodium group had more people with very poor kidney function.
Age 39 – 8% Died In 10 Years!
The mean age of the participants was only 39 at the beginning of the study, and yet 8% of them died and 4.5% developed end stage kidney disease in the following 10 years. This is way more than would occur in this age group.
Certainly this study group cannot be considered representative of a general population. The most likely reason for the worse survival in the low sodium group is that the sickest participants had been put on a low sodium diet. In other words, the high risk of serious illness and death prompted the participants to be on a low sodium diet, either because their doctor put them on it or because they did it themselves.
This is called reverse causality. They were not sick because they went on a low sodium diet. They went on a low sodium diet because they were sick (at high risk).
Another study in 2011 (3) showed a J-shaped relationship between urinary sodium and the combined outcome of stroke and CVD. The researchers combined non-fatal and fatal heart attacks, stroke and hospitalization for heart failure as an outcome measure. They showed a greater likelihood of these outcomes when sodium consumption was estimated to be less than 3000 mg a day.
One of the problems with this study is that it was a secondary look at a pair of studies designed for entirely different purposes, so that a great deal of needed information was not collected. As with the first study, it had a great many participants who had damage to their organs already. Almost half had already had a previous heart attack, more than 20% had a previous stroke, 40% had diabetes, and 70% had hypertension. Hardly representative of the general public.
The low sodium group had a higher percentage of people with prior heart disease, hypertension, and a sedentary lifestyle. They also had a higher percentage on diuretics (water pills).
A high percentage of this low sodium group was likely to have been put on a low sodium diet because of their health problems. Furthermore, they had distortion of the sodium in their urine because so many were using medications that affect sodium excretion into the urine. Another example of reverse causation.
These studies provide good examples of what to look out for when you hear about medical studies that you should be skeptical of. Reverse causation is a common mistake made when there is an association. Very few people can be made to think that evening cooling causes the sun to go down. But many can be fooled by claims of salt consumption below 3000 mg causing cardiovascular death. Only a few people will question the results. Especially when the study is presented by experts.
Find the posts with food tables listed with other posts under the List of Posts tab at the top of the page.
1. Sodium, blood pressure, and cardiovascular disease: further evidence supporting the American Heart Association sodium reduction recommendations. Whelton PK, Appel LJ, Sacco RL, Anderson CA, Antman EM, Campbell N, Dunbar SB, Frohlich ED, Hall JE, Jessup M, Labarthe DR, MacGregor GA, Sacks FM, Stamler J, Vafiadis DK, Van Horn LV. Circulation. 2012 Dec 11;126(24):2880-9. doi: 10.1161/CIR.0b013e318279acbf. Epub 2012 Nov 2.
2. The association between dietary sodium intake, ESRD, and all-cause mortality in patients with type 1 diabetes. Thomas MC, Moran J, Forsblom C, Harjutsalo V, Thorn L, Ahola A, Waden J, Tolonen N, Saraheimo M, Gordin D, Groop PH; FinnDiane Study Group. Diabetes Care. 2011;34:861– 866.
3. Urinary sodium and potassium excretion and risk of cardiovascular events. O’Donnell MJ, Yusuf S, Mente A, Gao P, Mann JF, Teo K, Sleight P, Sharma AM, Dans A, Probstfield J, Schmieder R. JAMA. 2011;306:2229–2238.