People Who Drink More Lose More

The empty tank, the deconditioning tax, and the two clocks — a twenty-five-year-old clinic rule, finally explained by a meter

Twenty-five years ago I ran a medical weight-loss program. I had a scale, a clipboard, and a sentence I said so often the staff could have said it with me.

People who drink more lose more.

I treated it as a clinic rule, not a scientific claim. I assumed it was mostly behavioural: someone busy drinking water is usually less busy snacking. That was good enough for practice. The scale kept agreeing with me, so I never went looking for the reason.

This summer I started watching myself much more closely. I measure blood ketones every day with a finger-prick meter, and I measure glucose the same way. For two ten-day stretches I also wore a continuous glucose monitor — a sensor on the arm that records a reading every five minutes, day and night. Two windows into the same body. And once both windows were open, a pattern showed up that was too consistent to ignore.

When my fluid tank ran low, my glucose rose above my own normal and my ketones fell and stayed low. When I fixed the water, the ketones came back the same day. A different pattern showed up after hard exercise my body was not conditioned for: glucose sitting higher for days, ketones suppressed for exactly as long, both lines returning together. Two patterns. Two clocks.

So this is not a theory piece. It is a field report. Part of it comes from published science. Part of it comes from repeatable self-measurement in a body living in a range the published science barely touches. Most ketone studies describe blood beta-hydroxybutyrate around 0.4 to 1.0 millimoles per litre. My floor is 4.0. This summer I have spent weeks above it and weeks below it, all on camera. That is not a small difference. It is the difference between reading about weather at sea level and trying to predict what happens on a mountain.

I want to walk through this slowly. The topic is strange. It is easy to make it dry, and it is easy to make it sound more certain than it is. I am going to try to do neither.

The words I'll use

You should not have to stop and look anything up.

BHB. Beta-hydroxybutyrate, the main ketone I measure in blood. A fuel the liver makes from fat. Also a signal molecule — it does not just burn, it tells cells how to behave. I will mostly say the ketone, or the number.

Ketosis. A state in which the liver is making more ketones and the body is using them as fuel. Above 0.5 on the meter you are in it. My floor is 4.0.

CGM. Continuous glucose monitor. Mine was a Dexcom G7, worn for two roughly ten-day sensor sessions in August. Think of it as a dashboard camera for glucose: it records the movie, not a snapshot. When the sensor is off, I am back to finger-sticks, and I will say which is which.

Glucose. Blood sugar. One of the body's main fuels, and the one the liver defends first.

Insulin. The hormone that moves glucose into tissue and stores fuel. In this story it matters for one reason: it is the strongest known brake on ketone production.

Cortisol. The body's alarm hormone. Among its many jobs, it tells the liver to make sugar.

Gluconeogenesis. The liver building new glucose from spare parts — amino acids, the glycerol in fat, lactate. The emergency generator in the basement.

Electrolytes. Charged minerals — sodium, potassium, magnesium — that control fluid balance, nerves and muscle.

Natriuresis. The kidneys letting sodium go in the urine. Water follows sodium, so this is one way a ketogenic diet quietly lowers body fluid volume.

Volume. The amount of fluid actually in the blood vessels. Not hydration in the water-bottle sense. The quantity in the pipes.

Orthostatic change. What the body does when you stand up. If the heart rate jumps or the pressure falls, the system is struggling with the transition — usually because there is less fluid to move.

Stroke volume. The amount of blood the heart pushes forward with one beat. A hand pump: each squeeze moves a certain amount of water. That amount is the stroke volume.

Pulse pressure. The gap between the top and bottom blood-pressure numbers. Think of a garden hose: when the tank behind the pump is full, the hose jumps with every stroke; when the tank is half empty, it barely twitches. A narrow gap is a clue about the tank. A clue, not a verdict — we will come back to that.

Vasopressin. The hormone that tells the kidneys to conserve water. The body turning the faucet down and the drain partly closed.

Deconditioning. A body that has lost its readiness for effort. Not weakness as a character flaw. Just what happens when the body has not been asked to do a job lately.

Now we can talk like normal people.

The morning I felt fine

September 3rd, mid-morning, at my desk, feeling completely normal. On a whim I took my blood pressure.

Eighty over sixty-eight.

A pulse pressure of twelve. The hose was barely twitching. My heart rate was eighty — high for me — because the pump was working harder to move less fluid. I checked again. Eighty-three over sixty-one. Eighty-eight over seventy-five. Same story each time.

I want to be clear about one thing. I felt fine. No drama, no collapse, no warning bell. If I had not owned a cuff I would have gone through the day, wondered that evening why my ketone number was flat, and blamed the food.

I drank water. Not a symbolic sip — a full glass, then another. Twenty minutes later: 102 over 68, heart rate 70. More water. 108 over 64, heart rate 62. The heart rate fell twenty beats and the pulse pressure went from twelve to forty-four, on water alone. And later that day the ketone number, which had been stuck, moved.

That was the third time this summer. There has since been a fourth. On September 8th the meter had read 3.4 for twenty-two hours straight no matter what I did. The cuff showed the same narrow gap, and my heart rate rose twenty-four beats when I stood up. I drank a litre and a half over the afternoon. The standing rise fell to fifteen within hours, and by evening the number was 4.6. Nothing else changed.

Every one of the four episodes had the same signature: a narrow gap on the cuff, a heart rate that jumped on standing, a ketone number that sat flat all day whatever I ate or did not eat — and then, hours after the water went in, the number climbed.

What a kitchen-table cuff can and cannot tell you

Now I need to be honest about my instrument, because I am a physician and this is where a physician should be most careful.

A home blood-pressure cuff is not a courtroom witness. It is a clue machine. Pulse pressure is pushed around by posture, cuff size, arterial stiffness, recent activity and ordinary measurement slop.[4] The idea that a narrow pulse pressure reflects a small stroke volume has real physiology behind it — it is why intensive-care physicians watch it during resuscitation — but the same physicians will tell you it is confounded by vascular tone and arterial compliance and should never be read alone.[3] You cannot diagnose the heart's output from a cuff on the kitchen table.

So I am not claiming that a narrow pulse pressure proves an empty tank. I am claiming something more modest and more defensible. It gave me a reason to look closer. Then the glucose and the ketone added a second and a third layer to the story, from independent instruments, and all three pointed the same way. That is stronger than the cuff by itself. It is still one man with three gadgets, and I will say so again before the end.

The workout that sent an invoice

The second pattern is slower and more expensive. That is why I call it a tax.

I am sixty. I have a history of knee pain that put me on a cane twice in two years. This summer I started training again, and on September 5th I ran for an hour without stopping — a sentence that would have been impossible in June. That story is told tissue by tissue in the knee piece. This one is about the meter.

Ten days before the run, on August 27th, I did a hard sixty-three-minute session after about ten days of doing almost nothing. That session fell inside one of my continuous-monitor windows, so for the day of and the day after I have the whole movie, and finger-sticks for the rest.

Here is what the movie showed, and I want to be exact because "spike" is the wrong word. My glucose did not spike. The highest single reading the sensor recorded in the entire fortnight was 121. What it did was sit higher than my normal for about three days: the sensor's daily average was 88 on the worst day against my usual high 70s, and the morning finger-sticks read 89, 106 and 105 on days I had not eaten. Not diabetic. Not even abnormal for most people. High for me. And for as long as it sat there, the ketone stayed below my floor — 2.6, 2.6, then 3.7, then back into the 5s.

The September run was on finger-sticks only, and it showed the same shape with a shorter tail: 3.7 before, 3.3 after, 3.1 the next morning, 2.6 that evening, 3.6 and climbing the morning after, 4.2 by the second afternoon.

That pairing is the headline. I do not just mean that both numbers changed. I mean they moved like two ends of a seesaw. When one stayed up, the other stayed down. When one came down, the other came back.

So there are really two patterns. One is the empty tank. It corrects within the day. The other is the deconditioning tax. It takes days. They may overlap — a hard session on a dry morning is exactly how the August 28th episode happened — but they do not run on the same clock. That difference in timing is the best clue in this whole piece.

Why a fast fix and a slow fix tell different stories

A correction that arrives within hours of a litre of water is not tissue repair. Cartilage does not heal in an afternoon. Muscle does not refill its glycogen stores in an afternoon. A knee does not settle an inflammatory cascade in an afternoon. But a hormone circuit can change course in an afternoon. A traffic light turns green in seconds. So the dry-day pattern looks like switching, not rebuilding.

The multi-day pattern after unaccustomed effort looks like a bill. More tissue demand. More repair signalling. More glucose being handled differently — one bout of exercise changes how muscle takes up glucose for many hours and sometimes up to two days afterward, and that was measured in a conditioned body.[8] A deconditioned body treats the same job as a bigger emergency. It recruits more help, patches more tissue, reorganises more traffic. A garage band booked into Carnegie Hall with no rehearsal: the music may happen, but the clean-up afterward will be ugly.

That is why I refuse to lump the two patterns together. They may touch each other. They are not the same animal.

What I think happens, one link at a time

I will label each link by how sure I am, because that is the only honest way to tell a mechanism story.

Link one: a low tank raises the alarm hormone. Established. When fluid volume drops, the body treats it as a threat. A 2008 study dehydrated trained men by two-and-a-half and five percent of body weight and measured their hormones through a workout: cortisol and adrenaline climbed with every step, and the hormones that build tissue fell.[5] A 2020 meta-analysis of twenty-one studies found cortisol significantly higher under dehydration, though the studies were messy and many mixed in heat, exercise or fasting.[6] And a 2025 study found the same thing in ordinary life: adults who habitually drink about a litre and a half a day mount a cortisol response to a stress test roughly fifty percent larger than adults who drink four litres, and the size of the response tracked how concentrated their urine was.[7] One caveat you should have: several of that study's authors work for a bottled-water company. The finding fits everything else; weigh it accordingly.

Link two: the alarm hormone tells the liver to make sugar. Established. Textbook physiology. Cortisol switches on gluconeogenesis and at the same time makes muscle and fat less willing to take sugar up. The result is more sugar in the blood.[9][10]

Link three: the real off-switch is insulin. Established for the switch; my observation for the chain. Here I have to correct something I have said before, including on camera. The liver does not choose between making sugar and making ketones the way a one-burner stove chooses between two pans. It is a utility plant with several output lines. It can run gluconeogenesis and ketone production at the same time, and in ordinary fasting both rise together.[9] So "cortisol went up, therefore ketones went down" is too crude a sentence.

The control point is insulin. Insulin is the red light for ketone production — this has been clear since Cahill's starvation studies half a century ago.[10] So the chain I think survives contact with both the physiology and my data runs: low tank or big unaccustomed stress → alarm hormones → more glucose produced or released → more insulin → ketone production held down. When the signal settles, insulin falls and the ketones come back.

The reason I put insulin at the centre is timing. In my log the ketone stays low exactly as long as the glucose stays up. If the ketone recovered while the glucose was still high, I would tell a different story. But that is not what I saw. I saw coupling. That does not prove every intermediate step. It does make the insulin-centred explanation stronger than the cortisol-only story I used to tell.

Link four: ketosis makes the tank easier to empty. Established plumbing; plausible consequence. There is a hidden plumbing issue here that people on ketogenic diets rarely hear explained. Low insulin does not only change sugar handling. It changes how the kidney handles sodium. Insulin tells the kidneys to hold on to sodium; when insulin falls, that signal weakens, the kidneys let more sodium go, and water follows.[11][12] That is natriuresis, and it is why people lose several pounds in the first week of any low-carbohydrate diet — much of it water. It is standard physiology, not controversy.

Why it matters here: a person living in deep ketosis may be walking around with less fluid reserve. Not sick. Not broken. Just carrying less cushion in the system. That means the distance between "fine" and "low enough to trigger trouble" may be shorter than most people realise — which could explain why a glass of water changes the whole picture so fast. Could. There is a difference between a plausible mechanism and a proven end-to-end human trial, and on that trial we are still waiting.

Link five: the fix is water, not salt. My observation, four times over. The first two episodes I read as salt loss and treated with salt. The third time I had salt on board and it did nothing until the water arrived. Salt without water is a half-treatment. You cannot fill a tank with a spoon.

Where the research pushes back

I do not want to hide the parts that argue against me. That would be sloppy. It would also be boring. A real argument should show the friction.

First. In ordinary fasting, cortisol, gluconeogenesis and ketone production all rise together.[9] So cortisol alone cannot explain a suppressed ketone. Agreed — that is why insulin is at the centre of my model and cortisol is upstream of it.

Second. If I am dehydrated, the same ketones are packed into less water, so the reading should look higher, not lower. That is a fair criticism and it runs directly against my story. My answer is not that it is wrong; it is that my number went the other way, repeatedly, after the water went in. Whatever concentration was doing, it was not the whole story — the hormone effect appears to be bigger than the concentration effect. But I cannot size either from a finger-prick. Somebody with a lab should.

Third. Point-of-care ketone meters are least precise above 5 millimoles per litre, which is exactly where my readings sit when things are going well, and hematocrit and hydration can distort them.[13] True. I treat the meter with suspicion. But the pattern is not coming from one gadget. The glucose tells an independent story at the same time, and noise wanders while this shape keeps repeating.

Fourth. Most published ketogenic trials never reach the level I aim for. The best-controlled diet trial of this year — forty-two adults with obesity, prediabetes and fatty liver, every meal cooked by a university kitchen, four percent carbohydrate, over ninety-five percent adherence — ran blood ketones between 0.4 and 0.8 all day, and the authors say plainly that was too low to be doing anything; they credit their impressive liver results to low insulin and high glucagon, not to the ketone.[1] That is a good trial and it says something I believe — what you measure decides what you learn. But it says nothing about the range I live in, and nothing about water. The literature is not useless here. It is incomplete. That is a more interesting complaint.

Fifth, and it is the sharpest. The same 2020 meta-analysis that found cortisol rising with dehydration found insulin unchanged.[6] If insulin is my traffic light, that is a problem. My answer is that those were short laboratory dehydrations in ordinary eaters, measured once — not a person at 4 millimoles per litre watching glucose and ketone move together across a day, in whom a small insulin change may matter far more than it does at sea level. But I cannot wave it away. It is the study I would run.

Why one person's data belong in this piece at all

A lot of health writing makes a mistake here. It assumes that if a randomised trial has not studied a situation, the situation does not exist. That is nonsense. Absence of a study is not absence of a phenomenon. It may simply mean nobody has looked.

My observations deserve to be here because they are direct measurements. The glucose is not inferred; it was recorded, by a sensor or a strip. The ketones are not imagined; they were measured, every day, and published the same day including the bad ones. The timing difference between a fluid correction and a deconditioned-exercise recovery is not vague; it is the whole point. And the pattern has repeated — four dehydration episodes, two workouts, the same two shapes.

No, that does not turn one person into a clinical trial. But it also does not make one person invisible. There is a place between "proven for everyone" and "not worth mentioning." That place is called an observation. Medicine starts there all the time: a careful physician sees something repeat, writes it down, and the argument begins. That is exactly where this sits, and I would rather say so than pretend to more.

What the scale knew twenty-five years ago

Now back to the original line. Was it always true? No. Was it true often enough to be a useful rule? Yes — for smaller reasons than I thought.

The strongest human evidence is modest. In a twelve-week trial, adults aged 55 to 75 on a reduced-calorie diet who drank two glasses of water before each meal lost 44 percent more weight than the group that did not — seven kilos against under five.[14] Most of that is appetite: water in the stomach before a meal means less food in it. A 2024 systematic review in JAMA Network Open looked at eighteen randomised trials of changing daily water intake and found most interventions did very little, with pre-meal water one of the few patterns that repeats.[15] A 2024 meta-analysis pooling eight trials in overweight adults found no clear overall effect on weight, body-mass index or waist — about a third of a kilo — though swapping water for sugar-sweetened drinks helped.[16]

And the famous claim that water boosts your metabolism by thirty percent? That comes from one small 2003 study,[17] and a sham-controlled 2015 reassessment found a change too small to matter.[18] A candle, not a furnace. Don't drink water to burn calories.

So the useful version of my old sentence is narrower: water helps when it replaces caloric drinks or takes up room before a meal. That is behaviour and mechanics, not a metabolic miracle. What twenty-five years of clipboards could not see — because nobody in the room was holding a ketone meter — is the part that matters most to anyone trying to live above a floor. An empty tank puts the liver on the wrong job.

What I think is certain, likely, and still unknown

Certain enough to stand on. Low-carbohydrate and ketogenic diets lower insulin and increase sodium and water loss through the kidneys.[11][12] Dehydration raises stress-hormone responses, especially under real physiological strain.[5][6][7] Insulin is the major suppressor of ketone production.[10] Exercise changes glucose handling and recovery signalling for hours to days.[8]

Likely, but not nailed shut. In a person living in deep ketosis, a low fluid tank raises glucose enough — through the alarm hormones and then insulin — to keep ketones suppressed until volume and electrolytes are corrected, and the correction takes hours, not days. My repeated observations support it. The physiology supports it. The direct human trial has not been done, and one meta-analysis found no insulin change in ordinary eaters.

More speculative, but worth saying out loud. Unconditioned hard exercise creates a prolonged stress-and-repair state that keeps glucose modestly elevated and ketones lower for about three days in a deep-ketosis body, and the bill shrinks as the body conditions. My data support it twice. Published work supports the pieces. No paper has studied this configuration. That is why I keep calling it uncharted — because it is.

Why this matters to me personally

I do not want this to read like a chemistry lecture that wandered away from the patient. I am the patient here — or at least the subject. I am sixty. I am a physician. I have a history of knee pain. I recently got through an hour-long run that would once have been off the table. I watched glucose sit higher for three days on a sensor while the ketone sat low. I watched a glass of water move a blood-pressure cuff twenty beats and, later that day, move the meter.

This is not a hobby argument about obscure molecules. It is me trying to understand the body I live in well enough to predict what comes next — which is the only kind of understanding that is any use to a patient.

Here is the sidewalk version. Sometimes the body is not failing. Sometimes it is sending an invoice. A dry system sends one kind of invoice; an unprepared workout sends another. Both are written in glucose first and ketones second. If that is true, learning to read the invoice matters — especially if you are trying to stay active, protect sore joints, and keep a metabolic experiment on the rails at the same time.

What I'd do with this, if I were you

Drink before you are thirsty. In deep ketosis thirst is a late signal, and later still. Every salt goes in with a full glass, never dry. Fasting means more water, not less — no food means no food-water. If you own a cuff, use it mid-morning on empty days and look at the gap between the two numbers, not just the top one; a narrow gap with a fast pulse is your tank talking before your thirst does. If you are short on sleep, know that you have opened the other alarm tap. And if you did something hard your body was not ready for, expect the bill to arrive in glucose first and ketones second and to take about three days to clear. Do not fix it by changing three things. Drink, salt with the drink, sleep, measure every morning, and watch it come back.

None of this is a diet. It is plumbing. It was true when I had a scale and a clipboard, and it is true now that I have a meter. The difference is that now I can tell you why — and where I still can't.

— Barry

I post my blood ketone number every day — including the bad ones — in a 90-second video: youtube.com/@neuraliftmd. Day one is there. So is every failure since.

If you want to do something like this with a physician watching, I'm taking on a small founding cohort. It starts with a conversation: calendly.com/hello-neuralift/30min

Educational only, not medical advice. Talk to your doctor before changing your diet, medications, or exercise.

Where all this comes from

1. Petersen MC, Smith GI, Farabi SS, et al. Effect of diet macronutrient content on the cardiometabolic response to weight loss: a randomized clinical trial. Cell Metabolism, published online August 27, 2026. doi:10.1016/j.cmet.2026.07.020. — 42 adults with obesity, prediabetes and fatty liver; ketogenic vs Mediterranean vs low-fat, all food supplied, identical ~10% weight loss. Liver fat −67% on keto vs −45%; prediabetes resolved in 50% vs 29% and 7%. Blood BHB on the ketogenic arm ≈ 0.4–0.8 mmol/L; the authors state it was too low to affect glucose production or insulin sensitivity. Verified from the full text.
2. My own record. Finger-prick BHB and glucose daily since June 17, 2026. Two ~10-day Dexcom G7 sensor sessions in August 2026 (the second covering Aug 17–27; 14-day report: no reading above 121, coefficient of variation 13.6%, daytime means 77–83 mg/dL). Dehydration episodes Aug 20, Aug 28, Sep 3 (80/68, HR 80 → 108/64, HR 62 on water), Sep 8 (3.4 flat 22 h → 4.6 by evening after 1.5 L; standing HR rise +24 → +15). Aug 27 workout: 63 min, average HR 126; next mornings 2.6 / 2.6 / 3.7 with glucose 89 / 106 / 105; sensor daily average 88 on Aug 27. Sep 5 run: 3.7 before, 3.3 after, 3.1 / 2.6 / 3.6 / 4.2 over the next two days. Personal data, not a trial.
3. Putowski Z, Teboul J-L, Castro R, et al. Can we reliably use pulse pressure as a surrogate for stroke volume? Physiological background and potential clinical implications for shock resuscitation. Critical Care 2025;29:249. doi:10.1186/s13054-025-05490-9. — Pulse pressure tracks stroke volume in physiology but is confounded by arterial compliance, vascular tone and wave amplification; a bedside clue, not a stand-alone measure. Verified.
4. Muntner P, Shimbo D, Carey RM, et al. Measurement of blood pressure in humans: a scientific statement from the American Heart Association. Hypertension 2019;73(5):e35–e66. doi:10.1161/HYP.0000000000000087. — Correct home technique and why single casual readings mislead. Verified.
5. Judelson DA, Maresh CM, Yamamoto LM, et al. Effect of hydration state on resistance exercise-induced endocrine markers of anabolism, catabolism, and metabolism. Journal of Applied Physiology 2008;105(3):816–824. doi:10.1152/japplphysiol.00862.2007. — Seven trained men at 0%, 2.5% and 5% dehydration; cortisol and norepinephrine rose with each step, testosterone response blunted. Verified.
6. Zaplatosch ME, Adams WM. The effect of acute hypohydration on indicators of glycemic regulation, appetite, metabolism and stress: a systematic review and meta-analysis. Nutrients 2020;12(9):2526. doi:10.3390/nu12092526. — 21 studies. Cortisol significantly elevated with hypohydration; insulin and ghrelin unaffected; large between-study variability. Source for link one AND for the fifth pushback. Verified.
7. Kashi DS, Hunter M, Edwards JP, Zemdegs J, Lourenço J, Mille AC, Perrier ET, Dolci A, Walsh NP. Habitual fluid intake and hydration status influence cortisol reactivity to acute psychosocial stress. Journal of Applied Physiology 2025;139(3):698–708. doi:10.1152/japplphysiol.00408.2025. — 16 habitual low drinkers (1.3 L/day) vs 16 high (4.4 L/day); Trier Social Stress Test; cortisol reactivity 6.2 vs 4.0 nmol/L, correlated with urine osmolality (r ≈ 0.7). Five co-authors are Danone Research employees. Verified (open-access PDF).
8. Mikines KJ, Sonne B, Farrell PA, Tronier B, Galbo H. Effect of physical exercise on sensitivity and responsiveness to insulin in humans. American Journal of Physiology 1988;254(3 Pt 1):E248–E259. — Seven untrained men, one hour of cycling; insulin-stimulated glucose uptake higher immediately and at 48 hours. Verified.
9. Herman R, Tršan J, Lipar L, Jensterle M, Janez A. Endocrine adaptations to prolonged fasting: from physiology, clinical uncertainties, translational challenges to healthspan implications. Nutrients 2025;17(24):3949. — Review: in prolonged fasting, cortisol, gluconeogenesis and ketogenesis rise together; catalogue of risks including hypercortisolaemia, orthostatic intolerance and electrolyte disturbance. Verified.
10. Cahill GF Jr. Fuel metabolism in starvation. Annual Review of Nutrition 2006;26:1–22. doi:10.1146/annurev.nutr.26.061505.111258. — The landmark review; insulin as the central control of ketone production. Verified.
11. Skartun O, Smith CR, Laupsa-Borge J, Dankel SN. Symptoms during initiation of a ketogenic diet: a scoping review of occurrence rates, mechanisms and relief strategies. Frontiers in Nutrition 2025;12:1538266. doi:10.3389/fnut.2025.1538266. — Mechanisms of keto-induction symptoms including natriuresis and fluid loss with falling insulin; rationale for electrolyte replacement. Verified.
12. Athinarayanan SJ, Roberts CGP, Vangala C, Shetty GK, McKenzie AL, Weimbs T, Volek JS. The case for a ketogenic diet in the management of kidney disease. BMJ Open Diabetes Research & Care 2024;12:e004101. doi:10.1136/bmjdrc-2024-004101. — Review of the natriuretic and diuretic effects of ketogenic diets and their implications for blood pressure and volume. Verified.
13. Dhatariya K. Blood ketones: measurement, interpretation, limitations, and utility in the management of diabetic ketoacidosis. Review of Diabetic Studies 2016;13(4):217–225. — What point-of-care ketone meters can and cannot tell you, including reduced precision at higher readings. Verified.
14. Dennis EA, Dengo AL, Comber DL, et al. Water consumption increases weight loss during a hypocaloric diet intervention in middle-aged and older adults. Obesity 2010;18(2):300–307. doi:10.1038/oby.2009.235. — 48 adults aged 55–75; 500 mL before meals; 44% greater loss (7.0 vs 4.8 kg) over 12 weeks. Verified.
15. Hakam N, et al. Outcomes in randomized clinical trials testing changes in daily water intake: a systematic review. JAMA Network Open 2024;7(11):e2447621. — 18 RCTs; most interventions had small or no effect; pre-meal water for weight loss and increased intake for kidney stones were the repeatable positives. Verified.
16. Chen Q-Y, Khil J, Keum N. Water intake and adiposity outcomes among overweight and obese individuals: a systematic review and meta-analysis of randomized controlled trials. Nutrients 2024;16(7):963. — Eight RCTs; pooled weight difference −0.33 kg (not significant); replacing sugar-sweetened beverages −0.81 kg. Verified.
17. Boschmann M, Steiniger J, Hille U, et al. Water-induced thermogenesis. Journal of Clinical Endocrinology & Metabolism 2003;88(12):6015–6019. — The source of the "30% metabolic boost" claim: ~30% rise in metabolic rate for about an hour after 500 mL, roughly 100 kJ. Verified.
18. Charrière N, Miles-Chan JL, Montani J-P, Dulloo AG. Water-induced thermogenesis and fat oxidation: a reassessment. Nutrition & Diabetes 2015;5:e190. doi:10.1038/nutd.2015.40. — Sham-controlled; effect much smaller than the 2003 estimate and not clinically meaningful. Verified.