Benefits
Exercise capacity after recovery from hard exercise
In an unregistered double-blind crossover trial in nine soccer players, a taste-matched deep ocean mineral drink taken during a 2 hour recovery raised later high-intensity running capacity, the stated primary outcome, compared with placebo. A taste-matched trial in 12 men found faster recovery of aerobic power, while trials in cyclists and triathletes found no endurance or race time difference.
Rehydration after dehydrating exercise
In a crossover trial in 17 adults who lost 3% of body mass through exercise in the heat, salivary osmolality, a hydration marker and the registered primary outcome, returned to baseline faster with a deep ocean mineral water (about 4 mg/L magnesium) than with spring water or a sports drink. Participants were not masked, and the water maker supplied the product.
Insulin sensitivity markers
In two Korean placebo-controlled trials of magnesium-rich deep sea water products in adults with prediabetes, fasting insulin and HOMA-IR were lower than on placebo after 8 and 12 weeks. Fasting and post-load glucose, the trials' main glucose measures, did not differ, and neither did HbA1c. These are blood markers, not functional outcomes.
Cholesterol, a blood marker
In a 6 week trial in 42 adults with high cholesterol, total cholesterol was lower with deep sea water (395 mg/L magnesium, 1,050 mL a day) than with plain water or with water given the same magnesium as magnesium chloride. A crossover trial also found lower total and LDL cholesterol than placebo, partly because levels rose during the placebo period, while a 78 person trial reported mainly within-group changes.
Gut environment markers
In a 12 week double-blind trial, healthy adults drank 1 L a day of deep sea water based drinking water or ordinary mineral water. Several stool short-chain fatty acids, among the stated primary markers, rose within the deep sea water group. Of those reporting constipation at the start, 16 of 17 improved on deep sea water versus 9 of 15 on mineral water, a questionnaire result.
Blood pressure and waist size
In an 8 week trial in adults with metabolic syndrome, 440 mL a day of deep sea water (about 163 mg magnesium) lowered waist size and systolic blood pressure within the group, but neither change differed from the control group. The control group drank no placebo water and the trial was not blinded.
Mechanism of action
Magnesium as the main mineral
Concentrated deep sea water is mostly magnesium, and the Korean insulin marker trials set their dose at 350 mg of magnesium a day, the Korean adult recommended intake. Magnesium is needed for normal insulin signaling, so the marker changes may reflect magnesium intake rather than anything unique to deep sea water.
Trace elements and the mineral mix
Trial authors have proposed that trace elements such as boron, lithium and rubidium, or the mineral mix as a whole, explain the exercise findings, but this has not been tested. In one cholesterol trial, deep sea water lowered total cholesterol more than water given the same magnesium as magnesium chloride, so magnesium alone may not account for that result.
Composition depends on the source
Products differ widely. Deep ocean mineral drinks used in trials ranged from about 4 mg/L to 813 mg/L of magnesium, and one undiluted concentrate held about 42,800 mg/L of magnesium. Results from one product cannot be assumed for another, and one exercise trial with similar findings used land well water, not seawater.
Clinical trials
Double-blind, counterbalanced crossover trial (randomization and trial registration not reported); about 1.4 L of a deep ocean mineral drink (diluted concentrate, about 171 mg/L magnesium, with 6% sucrose) or a taste-matched placebo during 2 hours of recovery after about 60 minutes of running; no external funding, and Pacific Deep Ocean Biotech paid only the open access fee (Higgins et al. 2019, Mar Drugs)
9 healthy recreational male soccer players, average age 22, in a thermoneutral room (20 °C).
High-intensity intermittent running capacity, the stated primary outcome, was greater after the mineral drink (1411 versus 905 seconds, P = 0.038); the authors estimated the minimum likely effect at about 25% after allowing for test variation and order effects. Blood lactate, blood glucose and urine osmolality did not differ.
Randomized, double-blind crossover trial; desalinated deep ocean water from 662 m off Hualien, Taiwan (140 mg/L magnesium) or reverse osmosis water, both taste-masked with the same additives, though the mineral water also held more sodium (119 versus 38 mg/L), at 1.5 times the body mass lost; water supplied by Taiwan Yes Deep Ocean Water; government grants (Hou et al. 2013, J Int Soc Sports Nutr)
12 healthy young men who ran at 30 °C until they lost 3% of body mass.
No primary outcome was named. Aerobic power (VO2max) returned to baseline within 4 hours with the mineral water while staying lower for 48 hours with placebo water, and leg power was about 10% above baseline at 4 and 24 hours. Creatine kinase and myoglobin rises were smaller; stress hormones did not differ.
Double-blind, placebo-controlled crossover trials in two age groups; desalinated deep ocean minerals from 618 m (146 mg/L magnesium) or taste-matched placebo before, during and after cycling at 75% of VO2max, 2 week washout; supported by Pacific Deep Ocean Biotech, the supplier (Wei et al. 2017, Front Physiol)
12 young men (average age 21) and 9 middle-aged men (average age 47).
Cycling time to exhaustion did not differ in either group (middle-aged men P = 0.08, effect size 0.09). Frontal brain hemoglobin measured by near infrared spectroscopy was higher, and in middle-aged men the post-exercise rise in neutrophil to lymphocyte ratio, an inflammation marker, was about 25% smaller.
Randomized counterbalanced crossover trial; rehydration with deep ocean mineral water (about 4.3 mg/L magnesium, 85 mg/L sodium), spring water or a carbohydrate sports drink equal to the fluid lost; registered NCT02486224 (masking of investigator and assessor only; terminated early); water supplied by Kona Deep; NIH and university funding (Harris et al. 2019, J Int Soc Sports Nutr)
17 adults (8 women, 9 men) dehydrated by 3% of body mass through exercise in the heat.
Salivary osmolality, the registered primary outcome, returned to baseline fastest with the deep ocean water (P < 0.0001). Rehydrating fluid affected leg extension peak torque recovery (P = 0.0157). A registered urine osmolality outcome was not reported, and the authors called the strength effect in need of further study.
Randomized, double-blind, placebo-controlled crossover trial; 440 mL a day of balanced deep sea water from 0.5 km off Goseong, Korea (813 mg/L magnesium, about 350 mg a day) or taste-matched placebo water for 8 weeks each, 4 week washout; registered NCT03956914 after completion; Korean government funding (Ham et al. 2020, Nutrients)
40 adults with prediabetes enrolled, 37 analyzed.
Blood glucose, the registered primary outcome, did not differ, nor did C-peptide or HbA1c. Fasting insulin (P = 0.042), HOMA-IR (P = 0.049), total cholesterol (P = 0.006) and LDL cholesterol (P = 0.003) were lower than on placebo, although none fell significantly within the deep sea water period, and cholesterol rose during the placebo period. Adverse events, laboratory tests and vital signs did not differ meaningfully.
Randomized, double-blind, placebo-controlled parallel trial; two 60 mL bottles a day of a drink made from spray-dried deep seawater mineral extract (from 1,500 m off Ulleungdo, Korea) or matched placebo for 12 weeks; the paper states a 350 mg magnesium target while its table lists 128.2 mg per bottle; registered KCT0008695 retrospectively; two authors employed by QBM Co., the maker; government funding (Ham et al. 2024, Medicina (Kaunas))
75 adults with impaired glucose tolerance analyzed (37 extract, 38 placebo).
Two hour glucose after a glucose load, the outcome used to size the trial, did not differ, nor did fasting glucose, glucose area under the curve or HbA1c. Fasting insulin, C-peptide, HOMA-IR (P = 0.0102 at 12 weeks), QUICKI and the Stumvoll index changed more favorably than on placebo. Adverse events were not reported.
Randomized three-arm trial; 1,050 mL a day for 6 weeks of reverse osmosis water, deep seawater (395 mg/L magnesium, hardness 1410 ppm) or water fortified with magnesium chloride (386 mg/L magnesium); blinding not stated in the abstract; Taiwanese university study (Fu et al. 2012, J Med Food)
42 adults with high blood cholesterol.
Total cholesterol fell progressively with deep seawater and was significantly lower than in the magnesium chloride water and plain water groups; LDL cholesterol and the lipid peroxidation marker TBARS also fell. Blood magnesium rose with deep seawater but stayed within the normal range in all groups.
Randomized, double-blind controlled trial; 1 L a day of refined deep-sea water (hardness 1000) or common soft mineral water for 12 weeks, labels masked; registered UMIN000022002; Japanese government funding; one author is from DyDo-T Beverage Co., a deep sea water beverage company, and the authors declare no conflict (Takeuchi et al. 2020, Nutrients)
98 healthy adults in Muroto, Japan (49 per group).
Five stool short-chain fatty acids rose and phenol and secretory IgA fell within the deep-sea water group, among the stated primary markers; in the adjusted analysis the water affected two of them (acetic and 3-methylbutanoic acids). Among people with constipation at the start, 16 of 17 improved versus 9 of 15 on mineral water (p < 0.05).