icon Москва
Это ваш регион? Да Выбрать другой


Consumption of low TDS water



With Review By

Dr. Lee t. Rozelle

and Dr. Ronald L. Wathen, M.D.



Since the beginning of time, water has been both praised for good health and blamed for human ills. We now know the real functions of water in the human body are to serve as a solvent and medium for the transport of nutrients and wastes to and from cells throughout the body, a regulator of temperature, a lubricator of joints and other tissues, and a participant in our body"s biochemical reactions. It is the H2O in water and not the dissolved and suspended minerals and other constituents that carry out these functions.

Low TDS water is defined in this paper as that containing between one and 100 milligrams per liter (mg/L) of total dissolved solids (TDS). This is typical of the water quality obtained from distillation, reverse osmosis, and deionization point-of-use water treatment of public or private water supplies that are generally available to consumers in the world.

Highly purified (distilled) water is believed by some to help "cure" arthritis by "washing out" excess calcium and other minerals from deposits in joints. Along with this reasoning, some people speculate that drinking highly purified water, treated by distillation, reverse osmosis, or deionization, "leaches" minerals from the body and thus causes mineral deficiencies with subsequent ill health effects.

An isolated report, a summary of Russian studies available through the World Health Organization, has recommended that fluid and electrolytes are better replaced with water containing a minimum of 100 mg/L of TDS. However, this may pertain more to situations in the human body during heavy exertion and sweating. It is the market for sports drinks which are formulated to help replace the sugar compounds, glucose in the blood and glycogen in the muscles that are burned in prolonged exercise. Sports drinks are formulated to help replace the sugar compounds - glucose in the blood, glycogen in the muscles - and electrolytes - salt, calcium, and potassium that keep cells in proper electrical balance - that may be burned and depleted after an hour or more of hard exercise. This situation does not have anything to do with low TDS or demineralized water for normal drinking and cooking purposes. Even in warm weather exercise, the greatest danger is that of dehydration, and the proper advice to ward it off is to drink lots of plain water.

The scope of this paper is limited to answering whether low TDS water contributes to the loss of minerals from body tissues, producing associated harmful side effects. The types of minerals — e.g., calcium versus sodium, or hard water versus soft water -- and the toxicity of minerals -- e.g., lead, cadmium, brackish, or saline waters — are not an issue in this report. Information on the body"s homeostasis mechanisms, community water supplies with natural TDS less than 50 mg/L, historic use of distilled water with less man three mg/L TDS on board Navy ships, the U.S. Environmental Protection Agency"s response to this issue, and other evidence are presented to demonstrate that the consumption of water with low levels of minerals is safe.


A review was conducted of the United States, Canadian, World Health Organization (WHO) and European Community (EC) drinking water standards. None of them has minimum limits or optimum levels of total dissolved solids. The U.S. recommended maximum level is 500 mg/L, the Canadian guideline suggests less than 1,000 mg/L, and the EC maximum admissible concentration (MAC) is 1,500 mg/L (for "dry residues").

The EC standards also list numbers for calcium (guide level of 100 mg/L) and magnesium (guide level of 30 mg/L., and MAC of 50 mg/L), and a minimum for hardness (minimum required concentration for softened water of 60 mg/L as Ca), and alkalinity (minimum required concentration for softened water of 30 mg/L HCCy). However, there is no health criteria documentation for these advisories. These levels are listed as aids to operation for water supply systems, i.e., suggested parameters for laying down a passivating film of scale in municipal distribution mains. Calcium, magnesium, hardness, and alkalinity conditions are not necessary for judging the safety of drinking water. It is understood that the association of European water suppliers, Bureau, is moving to exclude these parameters from the European legally enforceable limits.


A better understanding of the effect of low TDS water on the human body requires a basic understanding of the body"s mechanism in this respect. Following is a description of the relevant mechanism.

Homeostasis is the maintenance of static or constant conditions in the internal body environment. This natural process controls the mineral (ion) and the water concentrations in the body fluids within narrow limits inside and outside all the cells in all the organs and tissues of the body. The kidneys are most important in maintaining constant ion concentrations (including sodium, potassium, calcium, etc.) through elimination and reabsorption. In homeostasis, three body fluids are involved: plasma (approximately 3/5 of the blood volume); interstitial fluid (the fluid between cells); and intracellular (fluid inside the cells). The concentration of sodium ions is highest outside the cell and that of potassium ions is highest inside the cell. When the osmotic pressure is high on one side of the cell membrane (high concentration of ions) and low on the other side, water moves across the cell membrane from the dilute side toward the other side to equalize the osmotic pressure. This phenomenon is known as osmosis. (This is unlike reverse osmosis which occurs when outside pressure is applied to the concentrated side, pushing the water back to the dilute side.] The normal osmolality (concentration of ions) of all these fluids is about 300 milliosmoles per liter (mOsm/L (- 9,000 ppm).

Any changes from normal in ion concentration across the cell membrane is corrected in one minute or less because water moves quickly through cell membranes. Thus, small changes in osmolality from drinking purified water (0 to 100 mg/L TDS) are quickly brought to equilibrium.

The kidneys control the overall concentration of the constituents of body fluids. It filters about 180 liters (165 quarts) of water per day, but over 99% is reabsorbed and only 1.0 to 1.5 liters are eliminated as urine. If the osmolality of the fluid to be filtered by the kidney is lower than normal (low solute concentration - such as low TDS water) nervous and hormonal feedback mechanisms cause the kidney to excrete more water than normal and thus maintain the ion concentration in the body fluid to normal values. The opposite is true if the ion concentration of the fluid to be filtered is higher than normal. This kidney homeostatic mechanism keeps the body fluid osmolality normal. The osmolality of the fluid to be filtered by the kidney is controlled to ± 3% to maintain it at the normal level of 300 mOsm/L. The three basic hormonal and nervous control systems triggered by abnormal ion concentration in the body fluids to be filtered by the kidney are antidiuretic hormone (ADH) from the pituitary gland, aldosterone from the adrenal glands, and thirst (as osmolality rise of about 1% causes thirst).

Because of these kidney control mechanisms, drinking one liter of water would cause the urine output to increase about nine times after about 45 minutes (due to absorption of water in the gut) and continue for about two hours. Thus, the concentrations of solutes in the blood and other body fluids are quickly maintained by the kidney through homeostasis. These control mechanisms keep the sodium concentration at ± 7%. Calcium secretion is controlled by parathyroid hormone to ± a few percent in the extracellular body fluid.

Also, saliva increases the ion concentrations during water intake. The concentration of sodium chloride in saliva is typically 15 milliequivalents per liter (mEq/L) or 877 mg/L; that of potassium ion is about 30 mEq/L (1170 mg/L). As low TDS water is consumed, it is combined with saliva which increases the TDS before it reaches the gut to be absorbed, (e.g., each one milliliter of saliva can increase the TDS level in eight ounces of water consumed by about 10 mg/L).

Thus, based on the above highly credible and up-to-date textbook knowledge* it is evident that consumption by a healthy person of low TDS water alone cannot cause unhealthy systems. ["Healthy person" means free of disease, hormonal problems, etc., and not necessarily a healthy diet.] Of course, homeostasis is maintained by diet as are other body functions. If homeostasis is not maintained because of major diet deficiencies, disease, or hormonal dysfunction, consuming low TDS water would be a minor (if any) factor in any observed symptoms. It is apparent that disease, physiological dysfunction, or major nutritional deficiencies may cause a "leaching" problem, but not consuming one to two liters of low TDS water on a daily basis.

*Guyton, Arthur L., M.D. Textbook of Medical Psychology Eighth Edition, W.B. Saunders Company, Philadelphia (1991).


During the last 12 months, several literature searches have been undertaken to bring out studies, reports, reviews, and related information that may be of value to reach conclusions that would be scientifically supportable. These searches can be outlined as follows:

  • Request of expert review by U.S. EPA, Dr. Lee T. Rozelle, and Dr. Ronald L. Wathen of an annex (Appendix) attached to a report submitted to WHO prepared in 1980 by two Russians, G.I. Sidorenko and Y.A. Rachmanin, on the general subject of desalination.

  • Letters sent out to various experts and informed parties, including inquiries sent to the World Health Organization (WHO), U.S. EPA, and U.S. Department of the Navy.

  • A comprehensive search of the Medline medical literature database for articles related to low TDS water and homeostasis.

  • A search of AWWA"s WATERNET database for articles with key words describing low TDS waters and salt "leaching," etc. (No articles)

  • Review of published information regarding the levels of TDS in many public water supplies.

A review of the literature has shown that there is very little information published in western scientific literature that relates the consumption of low TDS water to physiological effects on the human body. A report "Guidelines on Health Aspects of Water Desalination" prepared by G.I. Sidorenko and Y.A. Rachmanin of Russia in 1980 and submitted to the World Health Organization contains an annex (Appendix), 61A pages long, reviewing work that has been reported in the Russian literature. The present literature search did not unearth any official translation of any of the articles cited in that annex. The annex concludes that consumption of water with less man 100 mg/L disturbs the body"s water/salt balance, promoting the release of sodium, potassium, chloride, and calcium ions from the body of animals or humans, imposing a stress on the mechanism of homeostasis, promoting changes in the gastrointestinal muscles and mucosa, and reducing the thirst quenching capacity of the water.

This annex has been intensively reviewed by many scientifically oriented individuals, including Dr. Lee Rozelle and Dr. Ronald L. Wathen.

Dr. Rozelle summarized his review:

The data for their conclusions, summarized in Annex 8 of the WHO unofficial guidelines, are not very convincing from a scientific viewpoint. The volume of water consumed per day was not indicated and the length of time of the experiment for the human "volunteers" was not indicated (one year for rats). The physiological changes reported apparently were based on rat and dog studies. For dogs, the same physiological changes were observed for water containing 50 mg/L and 1,000 mg/L.

The conclusion of a minimum TDS of 100 mg/L is confusing and thus not very convincing.

In the human studies, diuresis was observed particularly on the second day of the study (the increase in urine output reported to be 18%). The volume of water in the body from the "distillate" was reported to be 50 to 100% higher than the "other groups." The Russians also reported increased elimination of sodium, potassium, chloride, calcium, and magnesium in the urine, and the specific gravity was reduced. In the blood serum, the sodium was increased and the potassium decreased. Then it was reported that the "tendency" for similar changes were observed after consumption of 100 mg/L TDS and 1,000 mg/L TDS. The water intake (based on thirst) of various TDS waters was not clearly reported. In some cases it was difficult to determine if the data were from animals or humans.

In summary, the Russian studies, as reported in Annex 8 of the WHO document appeared not to be rigorously scientific.

Dr. Wathen reviewed it from a medical point of view and wrote the following:

The Annex Vin of the report is an alleged "review" of water and salt balance under the influence of a variety of water and salt intakes, in a variety of animals, including humans, exposed to a variety of conditions. The review is long on deductions, but very short on (re) presentation of solid data. Moreover, probably only a handful of references cited in this review (assuming they are cited correctly) may be from creditable scientific journals, that is, journals which demand proper scientific methodology and peer review of all work, prior to publication. Many of the cited articles may be from journals of "personal opinion": being versed only in English, it is impossible for me to establish the credibility of the cited work.

This review cited observations on the organoleptic features of water (i.e., consumer appreciation of taste, odor, and color qualities) to underscore precise, physiologic thirst slaking with specific levels of TDS — containing water in response to volume depletion. To begin with, the quoted electroencephalographic studies probably indicate only that a maximum number of receptor sites (taste buds) have to be recruited through stimulation to provide a maximum brain (alpha) wave response and that the TDS level in water providing the maximum response was between 200-600 mg/L of salt.

One would expect such a response; one might also imagine that receptor response (sensitivity) is considerably tempered by prior salt and mineral exposure for the individual. Organoleptic features are very, very unlikely to define whether a given water source is healthful on non-healthful or that the amount imbibed is appropriate to need. Moreover, taste receptor electrical activity, being unlikely to reveal preference by the consumer, means the consumer must be asked whether he or she prefers a given type of water. More often than not, preference reflects prior experience (learned behavior).

The review refers to exposure to desert and exercise conditions for humans and how water lost in sweating should be replaced, not with purified water but with salt water. Who would disagree with this conclusion? With Gatorade11, for example, selling to extremely large U.S. and world markets, one hardly needs to be reminded in this day and age that volume and salt losses encumbered with the sweating of heavy exercise are best replaced with a fluid whose constituents are more aligned with extracellular fluid in the human. GatoradeR, though billed a being "low sodium" on its label, is in fact rather high in salt content (both sodium and potassium) and it provides a rapid, convenient, and safe way to promptly reconstitute vascular volume after heavy exercise. Dizziness, even syncope (passing out), from volume depletion are thereby avoided and strength of the individual is better sustained, due both to volume replacement and the glucose contained within Gatorade.

GatoradeR is a water source to be taken only to replace severe fluid losses accompanying sweating; it generally averts the need for supplemental salt tablets. GatoradeR, in an opinion which may not be shared by the Gatorade Company (Chicago, IL), is not a source which should be used to replace normal fluid and electrolyte losses any more so than pure water should be used to accommodate severe volume losses. In the presence of salt accompanying usual dietary food intake (4-8 gm/day of sodium in U.S.), GatoradeR might conceivably lead to volume excess. A 64 oz. serving of Gatorade11 contains: 880 mg of sodium; 200 mg of potassium; and about 400 calories, as carbohydrate (112 grams). There is no magical solution; one should drink what"s appropriate to the circumstance.

Replacement of the large fluid losses accompanying heavy exercise or thermal exposure, therefore, has nothing to do with using purified water for normal drinking and cooking purposes, the latter being paired with normal food intake to meet salt and mineral needs. In normal day-to-day activity not associated with extremes in sweatloss, the salt and minerals accompanying normal food intake more than meet daily dietary needs of such elements, whether consuming potable water (e.g., TDS of up to 500 mg/L) or purified water (e.g., TDS 100 mg/L) or pure water (e.g., 0.111 mg/L TDS).

Salts and minerals are not "leached" from the human body; they are preferentially retained or excreted, either of these events occurring relative to whether or not one is surfeit in water or salt or both. In short, the human body is not a lead or copper pipe which "leaches" in the presence of purified water. The Annex Vin review is very misleading in this regard.

I also think it is incredible to suggest that, in the absence of abnormal water loading experiments, consumption of demineralized water will cause distortion of the mucosal cells lining the GI tract. Besides, in the normal human setting, such water is often combined with other elements (e.g., coffee, tea, fruit juices, soft drinks, etc.) which raises its TDS prior to consumption.

But, even if the TDS is not raised by some external means, through the additions of saliva, gastric secretions, and small intestine secretions beyond the stomach, there is an internal elevation in TDS of any dilute fluid one might drink. It would be my opinion that the adsorptive portion of the GI tract, that is, the small intestine, in the absence of extreme water loading, never sees a hypotonic solution sufficient to cause the mucosal cells to swell or to appear damaged. There is a tendency in this review to draw conclusions from observations in anhidrotic (non-sweating) laboratory animals (dogs, rats, and rabbits) and apply them to the very hidrotic (sweating) human. That is, the review often draws conclusions from fluid, electrolyte, and acid-base studies in "non-sweaters" and seems to transfer these conclusions to the "sweaters." It is imprudent physiologically, if not scientifically erroneous, to do so. I found the reported physiological data to be very confusing, often at right-angles to prior knowledge. One wonders if the responses to various levels of salt in water were confused when citing data from the actual papers used in the review.

I personally have never heard of osmorceptors being present in the gut which might serve to regulate fluid adsorption. And certainly, I have never heard of the liver being a repository of salt to be released to reconstitute salt levels in the vascular compartment. This statement, I think, is borderline preposterous. The literature cited here has been misconstrued or is comprised of essentially factitious observations. Salt in all body fluid compartments redistributes bidirectionally in attempts to off-set excesses or insufficiencies in extracellular fluid constituents, particularly in the vascular (blood volume) compartment.

The following responses have been received from different experts in many different fields:

WHO"s Dr. Galal-Gorchev states that WHO has "no information that such (low TDS) water would have and adverse effect on mineral balance".

U.S. EPA"s Dr. Edward V. Ohanian, Chief of Human Risk Assessment Branch wrote, Drinking water supplies a number of minerals that are important to human health. However, drinking water is normally a minor source of these minerals. Typically, the diet is the major source of these beneficial minerals. I am not aware of any data adequate to support the conclusion that water with low levels of minerals is unsafe.

The U.S. Navy has used distilled water with less than three ppm TDS aboard ship for more than 40 years. Surface ships while on shore take on water from shore sources, but it is common for submarines to provide nothing but purified water for months at a time, all with no reported Ш effects. This was confirmed with separate sources at the David Taylor Research Center in Annapolis, the Naval Sea Systems Command, the Bureau of Medicine and Surgery, and the Navy Environmental Health Center. Finally, the Surgeon General directed the Navy to address the subject formally in 1972. The conclusion was that drinking distilled water is not harmful.

The University of Illinois Health Sciences Library"s ONLINE SERVICES department was contracted to search the MEDLINE database for the several key word groups back to 1980. That search produced 18 articles, but only two relevant ones, both supporting the proposition that the physiologic mechanisms of homeostasis are more than adequate to permit routine drinking of purified water.

Article 1 described experiments in which dogs were given large amounts of distilled water to lower the blood osmotic pressure while monitoring very closely the pH and CO2 of arterial blood and the breathing rate. They found that the blood osmolality has a direct effect on breathing in addition to the known effects on kidney function. Thus, homeostasis appears to be controlled while intentionally consuming low TDS water.

Article 2 described experiments with rats which had had their pituitary glands removed ("hypophysectomized rats)", and which therefore should show some damage to the hormonal control system. They found that the renin-angiotensin-aldosterone system worked well even without a pituitary, suggesting an additional control system leading to the maintenance of homeostasis.

None of the 18 articles, including these two, showed any direct information about the subject matter of interest, but only that the homeostasis mechanism in living beings has significant abilities to adapt to different situations and inputs.


There are no known scientific data which clearly demonstrate that the consumption of low TDS water by humans will or will not lead to harmful effects on the human body. However, a number of field experiences can be cited which support the premise that the consumption of such water by humans does not cause such harmful effects. There are no known documented experiences which show that consuming low TDS water will create any long-term health effects.

The U.S. Navy has used distilled sea water for human consumption for approximately 40 years. TDS levels below 3 mg/L have been reported and consumption of this water for months at a time is common on submarines. No health problems have been reported by the Navy and they feel low TDS water is safe to drink.

The U.S. Army uses reverse osmosis units to provide drinking water for soldiers in the field. They do not consider low TDS water to be a problem and have no minimum standard.

The U.S. EPA conducted a project in San Ysidro, New Mexico in which the TDS was dropped from 800 mg/L to a range of 40 to 70 milligrams per liter. No health effects were observed during the one year test.

Possibly the largest field study of human consumption of low TDS water is within the United States where municipal systems are delivering water in this category. Millions of people currently consume such water, and this practice has gone on for decades. Exact data are difficult to obtain due to seasonal changes, use of blended water from multiple sources, and changes of sources. However, the following levels have been reported:

Boston, MA 64 mg/L

Portland, OR 23 mg/L

Baltimore, MD 89 mg/L

Lake Tahoe, CA 50-64 mg/L

San Francisco, CA 27-154 mg/L

Seattle, WA 34-47 mg/L

Denver, CO 39-216 mg/L

New York City, NY 41-283 mg/L

In addition to these areas, thousands of private wells, as well as numerous small municipal systems in the U.S., produce low TDS water. No known health effects or problems have been reported as a result of this widespread practice.

Thousands of TDS reduction devices have been sold residentially in the United States for decades. No reports of mineral depletion or health effects are known as a result of the consumption of this water.

In Plains, Texas, where the drinking water was brackish at 1500-2000 mg/L TDS, a change was made to desalinated water and the residents experienced temporary diarrhea. Because of the high initial TDS and because of blending, the treated water exceeded 100 mg/L TDS, which is above the definition of low TDS water established for this report.

In Montreal, a study compared the gastrointestinal disorders of two groups; one consumed tap water, while the other consumed reverse osmosis water. The group consuming the tap water had a higher incidence of GI infections. While neither of these experiences can be considered conclusive, no evidence of mineral leaching, the topic of this report, was reported in either case.

NASA has reported no Ш effects from the consumption of approximately .05 mg/L TDS water on board space craft. It appears that the possibility that this could have been a problem was never seriously considered by NASA.

In a field test in Boulder, Colorado with about 50 families, an experimental, zero discharge water system provided drinking water containing about.05 ppm TDS. No ill health effects were caused as a result of drinking this water.

In conclusion, the field experiences cited suggest that there are no long-term ill health effects, specifically the mineral leaching from human tissue, due to the consumption of low TDS water.


It has been concluded that the consumpdon of low TDS water, naturally occurring or received from a treatment process, does not result in harmful effects to the human body. This is based upon the following points:

  • No public health organization with authority over the drinking water quality anywhere in the world has enacted or even proposed a minimum requirement for total dissolved minerals in drinking water.

  • The human body"s own control mechanism (homeostasis) regulates the mineral content of the body fluids and the discharge of different types of ions from the body of normal health individuals drinking water with low or high mineral content.

  • Several types of scientific literature searches have found no harmful effects to the human body attributable to the consumption of low TDS water.

  • Review of the Soviet report has shown that the scientific methods used are questionable and the conclusions are either vague or unsupported by the data.

  • Many examples of real-world situations in which large populations have been and continue to be provided exclusively with low TDS water without any reported unusual or ill health effects, establishes the safety of consuming such waters by human beings.