LMNT: A Salt Company Recommending Higher Salt Intake, What Could go Wrong?
In this article, we delve into a series of claims put forth by LMNT within their collection of articles focusing on the scientific backing of their sodium intake recommendations. The intent here is to scrutinize the evidence provided in LMNT's articles and assess whether it adequately substantiates the claims they make.
The core assertion that stands as the linchpin of LMNT's argument is the recommendation of a sodium intake range between 4-6 grams. This is contrast to the recommendations of most, if not all, global health organizations. This article will meticulously examine the evidence presented by LMNT in support of this recommendation, as well as others, and determine whether the evidence they provide justifies their recommendations. By critically evaluating LMNT's stance and the scientific studies they reference, we aim to provide a balanced understanding of the intricacies surrounding sodium intake recommendations and the underlying evidence. Here are the articles that the claims were pulled from:
https://science.drinklmnt.com/electrolytes/lmnts-electrolyte-ratios-explained/
https://science.drinklmnt.com/electrolytes/the-whos-misguidance-on-sodium/
https://science.drinklmnt.com/electrolytes/the-fdas-misguidance-on-sodium/
Claim: 4-6g is sweet spot for sodium intake
I want to start with this one, because this is really the claim that is at the crux of LMNT’s entire argument. They are constantly arguing that the FDA/WHO is misguided in their attempts to lower population-level intake of sodium based on a handful of studies that all have the same fatal flaw.
LMNT Article Excerpt: “Athletes and folks on low-carb diets may need more sodium, but isn’t salt bad for your heart? The truth is, that’s not nearly as cut and dry as it’s been made out to be. Our 4–6 gram target comes from a 2011 JAMA study which found that 4–6 grams of sodium per day was the sweet spot for minimizing heart attack and stroke risk.”
The Study: “Urinary Sodium and Potassium Excretion and Risk of Cardiovascular Events” (2011)
Is this claim supported by the evidence provided?
Yes, but there’s more to it. This is a typical example of cherry picking a study of lesser quality when it fits your narrative, but ignoring the higher quality evidence that contradicts it. The below graph from the paper summarizes the results.
This study reported risk using a hazard ratio, which tells us the probability of a subject experiencing one of the outcomes given a specific sodium intake. A hazard ratio of 2, for example, means that they are twice as likely to experience one of the outcomes. As you can see, the hazard ratio of 1 (no increased risk) happens to occur at sodium excretion level of between 4-6g per day, with increased risk on either end of that range.
The main problem with this study and similar ones is the method used to collect sodium excretion from participants. They used a single sample of urine from each participant (spot urine collection) and used the Kawasaki formula to estimate 24-hour urinary sodium excretion. While spot urine is much easier to obtain both financially and logistically in studies, there are problems with using these formulas to predict CVD risk.
The Kawasaki formula uses age, sex, height, creatinine, and weight to estimate 24-hour sodium excretion, which are themselves predictors of hypertension (Naser et al., 2021). These issues lead to consistent overestimation of risk at “low” levels of sodium intake, due to an overestimation of sodium excretion. Another issue with spot urine collection, is it fails to account for interpersonal, day to day, fluctuation of sodium intake and output. In this study, estimates of daily sodium intake vary by over 800mg in half of subjects when multiple collections are used compared to a single one.
Preferred methods of estimating sodium intake are ones that involve both direct measurement and repeated measures (multiple samples taken of the same individual over time). For example, this study directly compared the mean of three to seven 24hr urinary sodium excretion measurements to estimated figures given by the Kawasaki formula when using spot urine measurements. They found that estimated values given by the Kawasaki formula overestimated actual measured values by 1297mg per day.
When we look at studies using these more robust and accurate measurement tools, the J curve turns into a linear one:
(He et al., 2018)
In conclusion, the only time we see this consistent J-shaped association between sodium and CVD risk is when estimated measures are used to predict sodium intake. But when sodium excretion is directly measured, the association becomes a linear one.
All this to say, the range of 4-6g of sodium intake is a deeply flawed recommendation to be giving to the general population.
Claim: Athletes lose up to 7g of sodium per day in hot climates
LMNT Article Excerpt: “Athletes lose up to 7 grams of sodium per day in hot climates. We’ve talked to trainers of professional athletes, and they often log up to a 10-gram sodium loss in a hard practice or game.”
The study: “Fluid and electrolyte needs for training, competition, and recovery” (2011)
Is this claim supported by the evidence provided?
Sort of. The specific figure of 7g of sodium loss actually comes from this paper which was cited in the above paper. There are two important distinctions in the LMNT claim that need to be clarified. First, the definition of “athlete”. I’ve been an athlete in Olympic weightlifting as well as Ironman triathlons, however, my sodium needs were much different in those two sports despite still being an “athlete”. The other important distinction is “hot climates”. So let’s take a look at how the paper defines those two things so we can understand the 7g sodium loss number more fully. The paper more so indicates that sodium losses of that magnitude are POSSIBLE in athletes exercising continuously for periods of 4-5 hours in hot temperatures (defined as > 30°C).
All in all, the LMNT claim is technically accurate but possibly misleading.
Claim: The FDA is unfairly attacking sodium and not sugar.
LMNT Article Excerpt: “And why did the FDA neglect even once to mention the dangers of sugar in its guidance to the food industry? A high-sugar diet (combined with inactivity) is a recipe for obesity and diabetes. The more people overeat, the more weight they gain, the higher their risk for metabolic disaster and heart disease. The science is clear on this point, and it’s no secret that sugar is both calorically dense and hyper-palatable.”
The Study: “Sodium Reduction” FDA publication
Is this claim supported by the evidence provided?
This is just a strange attempt at deflection. The article referenced by LMNT is specifically addressing sodium, so why would they talk sugar? There are separate statements published by the FDA on added sugars. In one of those statements, the FDA specifically states:
“The Dietary Guidelines for Americans recommends limiting calories from added sugars to less than 10 percent of total calories per day. For example, if you consume a 2,000 calorie daily diet, that would be 200 calories or 50 grams of added sugars per day. Consuming too much added sugars can make it difficult to meet nutrient needs while staying within calorie limits.”
Seems like the FDA and LMNT do agree on some things after all.
Claim: Reducing Salt Intake Doesn’t Help with Heart Failure
LMNT Article Excerpt: “a 2018 review (also in JAMA) found little evidence that reducing salt intake helps with heart failure.
The Study: Reduced Salt Intake for Heart Failure: A Systematic Review
Is this claim supported by the evidence provided?
Let’s take a look at what this study actually found. This was a systematic review of nine studies, which included 479 patients with heart failure. They were interested in two sets of outcomes, primary (cardiovascular-associated mortality, all-cause mortality, and adverse events, such as stroke and myocardial infarction) and secondary (hospitalization, length of inpatient stay, change in New York Heart Association [NYHA] functional class, adherence to dietary low-salt intake, and changes in blood pressure). The authors concluded that, of the nine studies, none of the included studies provided sufficient data on the primary outcomes of interest, and only four provided sufficient data for secondary outcomes. Of those four, two reported improvements in secondary outcomes from salt restriction, and two did not report improvement from salt restriction.
In summary, this review found inconclusive evidence one way or another regarding the effectiveness of salt reduction in the treatment for heart failure. There are two problems with the use of this study to support the 4-6g of sodium recommendation. First, this looked at a very specific population (heart failure patients) so you can’t just take the findings from something like this and apply it to the general population. For more on the importance of specifying a population for a research question, read here. Second, inconclusive evidence one way or another of a salt reduction intervention is a far cry from actively encouraging salt intakes far exceeding current recommendations.
Claim: Higher sodium intake is not associated with high blood pressure.
LMNT Article Excerpt: “The Intersalt Study looked at over 10,000 people across 48 global populations and found NO correlation between salt intake and the prevalence of hypertension. If salty diets raised blood pressure, it would show up here.”
Is this claim supported by the evidence provided?
First things first, when someone presents a 35-year old study as one of the foundational pieces of evidence for their claim, a little red flag should go up. Especially on a topic as ubiquitous as the link between sodium and blood pressure. There have been hundreds, if not thousands of studies on this topic since then. Regardless, let’s take a look…
The only explanation I can think of that would lead someone to connect this study to this claim is if someone briefly glanced at the abstract, found the sentence they were looking for to justify their conclusion and now present this study as “evidence”. Yes, this sentence does exist in the abstract:
“Across the other 48 centres sodium was significantly related to the slope of blood pressure with age but not to median blood pressure or prevalence of high blood pressure.”
However, that does not tell the whole story. Firstly, this study suffers from the same limitation as the previously discussed study, in that they used single 24-hour excretion measurements to estimate an individual’s sodium intake (the authors list this as a limitation themselves). But let’s press on anyways. I strongly recommend following the link above to this study and reading the discussion section on your own, but I will attempt to consolidate the main points without getting too deep into the statistical methods used, although they play a large role in how these results have been misinterpreted.
To summarize, this study looked at the relationship between sodium excretion and blood pressure across 52 medical centers. Importantly, they controlled for other known factors of hypertension such as BMI, and alcohol consumption. There were four centers with abnormally low sodium excretion and blood pressure which affected the strength of the relationships, and so the authors presented all analysis with and without those four centers included. There are two sets of analyses that are going to be important here: 1. Within center analysis 2. Across center analysis.
In the within center analysis, 33 of the centers found positive associations between sodium excretion and systolic blood pressure, 8 of which were statistically significant. Now, it is important here that I mention that something not being “statistically significant” does not mean you should completely disregard it. When we talk about statistical significance, there is one factor that influence whether something crosses the threshold for significance (usually p <. 05) that is relevant here. The relationship between sample size and expected effect size. As sample sizes increase, the ability to detect smaller and smaller effects becomes possible. However, the opposite is also true. There could be a small, but clinically meaningful effect present, but if a sample size is too small it will not be detected (resulting in a p >.05). This is important to mention because in the within center analyses sample sizes were usually under 200 people, and we expect changes in blood pressure to be numerically pretty small. It’s also important to realize the a p value of .05 is not some magical threshold, it is simply a useful gauge to delineate significant or non-significant results.
In the across center analysis, regression coefficients from each of the centers were pooled together and analyzed holistically. In this analysis, sodium excretion and slope of both systolic and diastolic pressure with age(the rate at which blood pressure increases with age) remained positive. However, as LMNT will point out, the overall relationship between sodium excretion and median blood pressure was not significant. But instead of taking that one sentence at face value, let’s hear what the authors have to say about why we may see this relationship in the within center analysis and not the across center analysis:
“Several factors may have contributed to these inconsistencies. Firstly, the statistical power to detect significant associations in the cross centre analyses was low once the four populations with low sodium values were excluded, given a lower than expected upper limit of sodium intake. Secondly, multiple factors might affect blood pressure across populations-for example, climate, physical activity, level of acculturation-but are less likely to confound within centre comparisons; inability to control for these effects might bias association in cross centre (ecological) analyses. For one important variable-slope of blood pressure with age- significant positive associations with sodium were found across centres, consistent with the within centre findings. Systematic measurement biases might distort comparisons of medians across centres, but, being similar across age groups and individual subjects within a centre, would be less likely to bias values of the slope of blood pressure with age or within centre comparisons”
In fact, the authors of the study themselves would not agree with LMNT’s interpretation of the results. Here are the takeaways that the authors provide from the data:
“The following conclusions may be drawn from the lntersalt data:
(1) Within centres sodium excretion was significantly related to blood pressure in individual subjects, and at least in part this relation was independent of body mass index and alcohol intake. Sodium excretion was also significantly related across centres to the slope of blood pressure with age. Thus lower average sodium intake might have a favourable influence on blood pressure, on change of blood pressure with age, and hence on cardiovascular mortality.
(2) The four Intersalt populations with low sodium excretion had low median blood pressures, low prevalence of hypertension, and either a decrease or only a small increase of blood pressure with age.
(3) Cross centre analyses of sodium excretion and median blood pressure and of sodium excretion and the prevalence of hypertension showed a positive association when all 52 centres were included but not when the four populations with low sodium values were excluded. The inconsistency with the findings in individual subjects might reflect the role of confounding variables that differed widely across centres but less so within centres.
(4) Potassium excretion was negatively and independently associated with blood pressure of individual subjects within centres after adjustment for sodium excretion, body mass index, and alcohol intake.
(5) The relation of the urinary sodium to potassium ratio to blood pressure in individual subjects followed a pattern similar to that for sodium but more strongly and consistently.
(6) Body mass index and high alcohol intake were strongly, positively, and independently associated with blood pressure in individual subjects.”
To recap, using this study to claim unambiguously that sodium has no relationship with blood pressure is just not accurate.
Main Takeaways
When governmental agencies develop population-level guidelines, they must consider what intakes would present the lowest risk for the greatest number of people. With that in mind, the data does support the recommendation for most people to limit sodium to less than 2300mg. The assertion from LMNT and others that lower amounts of intake INCREASE risk, is solely based on faulty sodium excretion estimation data as we’ve seen. This invalidates the broad 4-6g recommendation that they promote. Nothing against LMNT specifically, they have a product that is very useful for the right person. However, a broad recommendation to the general population to consume 4-6g of sodium per day for “optimal” health outcomes does not have strong evidence to support it.
Who Actually Needs More Sodium?
Ok, now that we have established that minimizing sodium intake in accordance with FDA and WHO recommendations is a good idea for MOST people, it is only fair that we talk about people who SHOULDN’T abide by these recommendations and may need sodium above and beyond those amounts. To be clear, if you goal is just general health and wellness, you are better off limiting sodium in accordance with the FDA guidelines in most cases. However, as athletes sodium intake can have important implications for performance if not properly addressed.
Due to the huge variation between individuals in regard to sweat rate and saltiness of sweat, it is difficult to give definitive recommendations that fit all athlete types. But, using the latest research, we at BIN have created a tool that lets you estimate the amount of fluid and sodium loss based on how your weight changes before and after exercise.
Based on this, you can have a general idea of how much sodium and fluid you need to replenish based on a certain activity.
Download the tool here:
Instructions For Using the Tool
**Only change values of ORANGE cells**
Sweat Loss Section
Step one- Void your bladder, weigh yourself and record it to the nearest 0.1kg, preferably without clothes on.
Step two- Complete your workout. If you ingest any fluid during your workout, be sure to measure exactly how much.
Step three- Immediately after your workout, towel yourself dry and weigh yourself again to the nearest 0.1kg
Step four- Enter your pre and post measurements and any fluid ingested into the spreadsheet.
Step five- The green cells will now display your estimated fluid and sodium losses during that exercise, which you may consider replenishing in excess of the recommended 2300mg/day. You can also take these numbers and annotate them in the log on the second tab (more instructions below).
Sweat Rate Section
Enter the duration of the exercise session (in hours). With this, you will be able to estimate your fluid and sodium losses during similar activities of varying durations. I suggest using the second tab of the spreadsheet (hydration log) to keep a detailed account of your fluid/sodium losses under various conditions. This way, the next time you have a similar session planned to one that you logged previously, you can pre-emptively plan your hydration strategy for that day.
Planned Hydration Calculator
In this section, you can take the expected temperature and duration of an event/workout and determine an estimate fluid/sodium loss from that event/workout. These outputs are based off your specific entries in the log. The more entries you have, the more accurate the estimations will become. Try to get as many measurements as possible at a variety of temperatures to really increase the accuracy. Remember to put the sweat RATE measurements from each workout, not total losses.
Simply enter the expected air temperature of the event/workout and the expected duration in hours. The “Expected Total Fluid Loss” and “Total Expected Sodium Loss” are the amount of water and sodium that you want to replenish either after the workout, or by adding those numbers to your normal intake for the day. Here is an example of a filled out log and planned event:
If you have questions or feedback about this tool, feel free to reach out to me at chris@blackironnutrition.com.